US6067300A - Method and apparatus for optimizing the transfer of data packets between local area networks - Google Patents

Method and apparatus for optimizing the transfer of data packets between local area networks Download PDF

Info

Publication number
US6067300A
US6067300A US09/096,110 US9611098A US6067300A US 6067300 A US6067300 A US 6067300A US 9611098 A US9611098 A US 9611098A US 6067300 A US6067300 A US 6067300A
Authority
US
United States
Prior art keywords
buffer
packet
mac
receive
data
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Lifetime
Application number
US09/096,110
Inventor
Robert J. Baumert
Anthony W. Seaman
Sherre M. Staves
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Extreme Networks Inc
Original Assignee
Cabletron Systems Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Cabletron Systems Inc filed Critical Cabletron Systems Inc
Assigned to NETVANTAGE reassignment NETVANTAGE ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: BAUMERT, ROBERT J., SEAMAN, ANTHONY W., STAVES, SHERRE M.
Priority to US09/096,110 priority Critical patent/US6067300A/en
Priority to EP99927458A priority patent/EP1086552B1/en
Priority to CA002328220A priority patent/CA2328220C/en
Priority to PCT/US1999/013166 priority patent/WO1999065193A1/en
Priority to AU44356/99A priority patent/AU767085B2/en
Priority to DE69941697T priority patent/DE69941697D1/en
Assigned to CABLETRON SYSTEMS, INC. reassignment CABLETRON SYSTEMS, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: NETVANTAGE
Publication of US6067300A publication Critical patent/US6067300A/en
Application granted granted Critical
Assigned to ENTERASYS NETWORKS, INC. reassignment ENTERASYS NETWORKS, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: CABLETRON SYSTEMS, INC.
Assigned to OBSIDIAN, LLC, WELLS FARGO FOOTHILL, INC. reassignment OBSIDIAN, LLC SECURITY AGREEMENT Assignors: ENTERASYS NETWORKS, INC.
Assigned to WELLS FARGO TRUST CORPORATION LIMITED, AS SECURITY AGENT reassignment WELLS FARGO TRUST CORPORATION LIMITED, AS SECURITY AGENT GRANT OF SECURITY INTEREST IN U.S. PATENTS Assignors: ENTERASYS NETWORKS INC.
Assigned to ENTERASYS NETWORKS, INC. reassignment ENTERASYS NETWORKS, INC. RELEASE AND REASSIGNMENT OF PATENTS AND PATENT APPLICATIONS AT REEL/FRAME NO. 17656/0552 Assignors: ENTERPRISE COMMUNICATIONS FUNDING GMBH, AS SUCCESSOR IN INTEREST TO OBSIDIAN, LLC, WELLS FARGO CAPITAL FINANCE, INC. (FORMERLY KNOWN AS WELLS FARGO FOOTHILL, INC.)
Assigned to ENTERASYS NETWORKS INC. reassignment ENTERASYS NETWORKS INC. TERMINATION AND RELEASE OF SECURITY INTEREST IN PATENTS AT REEL/FRAME NO. 25339/0875 Assignors: WELLS FARGO TRUST CORPORATION LIMITED
Assigned to SILICON VALLEY BANK reassignment SILICON VALLEY BANK SECURITY AGREEMENT Assignors: ENTERASYS NETWORKS, INC.
Assigned to EXTREME NETWORKS, INC. reassignment EXTREME NETWORKS, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: ENTERASYS NETWORKS, INC.
Assigned to EXTREME NETWORKS, INC. reassignment EXTREME NETWORKS, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: ENTERASYS NETWORKS, INC.
Assigned to SILICON VALLEY BANK reassignment SILICON VALLEY BANK AMENDED AND RESTATED PATENT AND TRADEMARK SECURITY AGREEMENT Assignors: EXTREME NETWORKS, INC.
Assigned to SILICON VALLEY BANK reassignment SILICON VALLEY BANK SECOND AMENDED AND RESTATED PATENT AND TRADEMARK SECURITY AGREEMENT Assignors: EXTREME NETWORKS, INC.
Assigned to SILICON VALLEY BANK reassignment SILICON VALLEY BANK THIRD AMENDED AND RESTATED PATENT AND TRADEMARK SECURITY AGREEMENT Assignors: EXTREME NETWORKS, INC.
Assigned to ENTERASYS NETWORKS, INC. reassignment ENTERASYS NETWORKS, INC. RELEASE BY SECURED PARTY (SEE DOCUMENT FOR DETAILS). Assignors: SILICON VALLEY BANK
Assigned to BANK OF MONTREAL reassignment BANK OF MONTREAL SECURITY INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: EXTREME NETWORKS, INC.
Assigned to EXTREME NETWORKS, INC. reassignment EXTREME NETWORKS, INC. RELEASE BY SECURED PARTY (SEE DOCUMENT FOR DETAILS). Assignors: SILICON VALLEY BANK
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/90Buffering arrangements
    • H04L49/9047Buffering arrangements including multiple buffers, e.g. buffer pools
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/35Switches specially adapted for specific applications
    • H04L49/351Switches specially adapted for specific applications for local area network [LAN], e.g. Ethernet switches
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/90Buffering arrangements
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/90Buffering arrangements
    • H04L49/901Buffering arrangements using storage descriptor, e.g. read or write pointers
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/90Buffering arrangements
    • H04L49/9042Separate storage for different parts of the packet, e.g. header and payload
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/90Buffering arrangements
    • H04L49/9057Arrangements for supporting packet reassembly or resequencing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/90Buffering arrangements
    • H04L49/9084Reactions to storage capacity overflow
    • H04L49/9089Reactions to storage capacity overflow replacing packets in a storage arrangement, e.g. pushout
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/30Peripheral units, e.g. input or output ports
    • H04L49/3018Input queuing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/30Peripheral units, e.g. input or output ports
    • H04L49/3027Output queuing

Definitions

  • the present invention relates generally to a method and apparatus for transferring data packets between a plurality of local area networks (LANs) and more particularly to a switch architecture capable of high speed packet transfer between multiple data ports (e.g., 28 Ethernet ports comprised of 10/100 and/or 10 Mbps ports)
  • LANs local area networks
  • switch architecture capable of high speed packet transfer between multiple data ports (e.g., 28 Ethernet ports comprised of 10/100 and/or 10 Mbps ports)
  • U.S. Patents directed to apparatus for switching or bridging LANs include U.S. Pat. Nos. 4,627,052; 4,715,030; 4,737,953; 4,922,503; 5,136,580, and 5,633,865.
  • the present invention relates to a method and apparatus for optimizing the transfer of data packets between a plurality of local area networks (LANs).
  • Apparatus of the present invention are comprised of multiple independent controllers, e.g., a receive controller and a transmit controller, which share common resources including a first memory (a packet memory) which stores the data packets, a second memory (a descriptor memory) which stores pointers to the stored data packets, and buffered data paths (preferably using FIFO buffers).
  • the independent controllers operate essentially concurrently for most tasks while interleaving their use of the shared resources. Consequently, embodiments of the present invention can simultaneously receive and transmit data across multiple LAN data ports (e.g., 28 Ethernet ports comprised of 10/100 and/or 10 Mbps ports).
  • a system for selectively transferring data packets between a plurality of local area networks (LANs), wherein the data packets identify destination and source addresses comprises: (1) a plurality of media access controllers (MACs) each configured for receiving and transmitting data packets to and/or from a different LAN, the MACS including at least a first MAC associated with a first LAN for receiving data packets from the first LAN and a second MAC associated with a second LAN for transmitting data packets to the second LAN, (2) a MAC receive buffer for periodically receiving at least a portion of data packets from the MACs, (3) a MAC transmit buffer capable of operating concurrently with said MAC receive buffer for periodically transmitting at least a portion of data packets directed to the MACs, (4) a shared packet memory for storing a plurality of data packets, (5) a packet memory input buffer for periodically receiving and buffering at least a portion of data packets from the packet memory, (6) a packet memory output buffer capable of operating
  • the system additionally includes: (1) receive session registers for storing descriptors generated by the receive controller before the descriptors are stored in the descriptor memory and (2) transmit session registers for storing descriptors retrieved from the descriptor memory, wherein the transmit controller is responsive to the retrieved descriptors within the transmit registers for selecting data packets from the packet memory for transmission to the MACs.
  • the aforementioned buffers are organized as first in first out buffers (FIFOs) capable of inputting data at a first rate and outputting data at a second rate.
  • FIFOs first in first out buffers
  • the aforementioned buffers and the receive and transmit controllers are all implemented within a single integrated circuit package.
  • FIG. 1 comprises a top level functional block diagram of a preferred system for optimizing the transfer of data packets between a plurality of local area networks (LANs) using a switch controller;
  • LANs local area networks
  • FIG. 2 comprises a functional block diagram of the switch controller of FIG. 1;
  • FIG. 3 shows the basic format of an exemplary data packet transferred by the present invention
  • FIG. 4 shows a simplified block diagram of the data components allocated in the packet memory and descriptor memory of FIG. 1;
  • FIG. 5 shows an exemplary functional block diagram of the elements used in performing the receive task of the present invention.
  • FIG. 6 shows an exemplary flow chart of the receive task of the present invention which utilizes multiple concurrent state machines
  • FIG. 7 shows an exemplary functional block diagram of the elements used in performing the transmit task of the present invention.
  • FIG. 8 shows an exemplary flow chart of the transmit task of the present invention which utilizes multiple concurrent state machines.
  • FIG. 9 shows an exemplary timeline showing multiple receive and transmit tasks concurrently using different shared resources to concurrently receive and transmit data packets on a plurality of LANs.
  • the present invention relates to a method and apparatus for optimizing the transfer of data packets between a plurality of local area networks (LANs).
  • LANs local area networks
  • FIG. 1 shows a functional block diagram of a preferred system 10, commonly known as a switch, for optimizing the transfer of data packets between a plurality of local area networks (LANs) e.g., 12, 14, and 16, according to data contained within each data packet (described further below in reference to FIG. 3).
  • LANs local area networks
  • the system 10 is primarily comprised of (1) a plurality of media access controllers (MACs) 18 each corresponding to a different LAN 12, 14, 16 for bidirectionally transferring data packets, (2) a shared packet data memory 20 (preferably RAM and referred to as packet memory) for storing data packets prior to transferring to another LAN, (3) a shared packet descriptor memory 22 (preferably RAM and referred to as descriptor memory) for storing pointers identifying the location of the data packets in the packet memory 20, and (4) a switch controller 23 defining a receive data path 24 and a transmit data path 25 for concurrently transferring data in both directions between the MACs 18 and the packet memory 20.
  • MACs media access controllers
  • the switch controller 23 periodically shares resources including the packet memory 20 and the descriptor memory 22.
  • periodic is intended to refer to events occurring at regular or irregular intervals.
  • the independent controllers within the switch controller 23 operate essentially concurrently for most tasks while interleaving their use of the shared resources. Consequently, embodiments of the present invention can simultaneously receive and transmit data across multiple LAN data ports, e.g., LAN 1 12, LAN 2 14, LAN N 16, etc., via a plurality of media access controllers 18.
  • Table I shows a simplified example of multiple receive tasks, i.e., a first data packet RX 1 being received from a first LAN 1 (column 2) and a second data packet RX 2 being received from a second LAN 2 (column 3), being processed concurrently with a third data packet TX 3 being transmitted on a third LAN 3 (column 4) in a time distributed manner during each time slot T X .
  • the time slots T X can have variable durations for each task, e.g., as needed for a particular subtask.
  • the transmit task of column 4 may overlap different portions of the receive tasks and may have unequal time slot durations.
  • the media access controllers 18 preferably interface to the LANs via a PHY, i.e., a physical interface layer, (not shown).
  • a PHY i.e., a physical interface layer
  • 28 Ethernet ports comprised of 10/100 and/or 10 Mbps ports are simultaneously supported.
  • FIG. 2 shows the switch controller 23 of FIG. 1 primarily comprised of (1) a MAC receive (RCV) buffer 30 for periodically receiving and buffering at least a portion of data packets from the MACs 18 via a bidirectional MAC interface bus 31, (2) a MAC transmit (XMIT) buffer 32 capable of operating concurrently with the MAC receive buffer 30 for periodically buffering and transmitting at least a portion of data packets directed to the MACs 18 via the MAC interface bus 31, (3) a packet memory input buffer 34 for periodically receiving and buffering at least a portion of data packets from the packet memory 20 via a bidirectional packet memory interface bus 35, (4) a packet memory output buffer 36 capable of operating concurrently with the packet memory input buffer 34 for periodically buffering and transmitting at least a portion of data packets to the packet memory 20 via the packet memory interface bus 35, (5) an address table 38 for correlating the destination address of each data packet with one of the MACs 18 and its corresponding LAN, (6) a receive (RCV) controller 40 for causing a data packet from
  • embodiments of the invention preferably include a CPU interface bus 44 to a CPU 46 which can monitor and/or alter the flow of data within the switch controller 23.
  • a CPU transmit (XMIT) buffer 48 and a CPU receive (RCV) buffer 50 which perform a similar function to that of the MAC transmit 32 and receive 30 buffers.
  • XMIT CPU transmit
  • RCV CPU receive
  • a multiplexer 51 under control of the transmit 42 and receive 40 controllers is used to direct data between the packet memory 20 to either the MACs 18 or the CPU 46 via its associated buffered interfaces 31, 35, and 44.
  • the MAC receive buffer 30 is essentially coupled to the packet memory output buffer 36 via path 52 to form the receive data path 24.
  • the MAC transmit buffer 32 is essentially coupled to the packet memory input buffer 34 via path 53 to form the transmit data path 25.
  • embodiments of the present invention are capable of concurrently receiving and transmitting data packets on a plurality of LANs by using a plurality of controllers (e.g., 40-42) which concurrently process portions of data packets.
  • Table II shows an expanded view of the simplified example previously shown in Table I corresponding to the preferred implementation of the receive 24 and transmit 25 data paths shown in FIG. 2.
  • FIG. 3 shows the basic structure of an exemplary received data packet 54 (e.g., for an Ethernet data transfer protocol) comprised of (1) a preamble 56 for designating the beginning of the data packet 58 (BOP), (2) a destination address (DA) 60 for identifying a destination device present on a destination LAN, e.g., LAN 2 14, (3) a source address 62 (SA) for identifying an originating device present on a source LAN, e.g., LAN 1 12, (4) a data field 64 containing the information content of the data packet 54, and (5) a CRC field 66 at the end of the data packet (EOP 68) for performing an integrity check on the contents of the data packet 54.
  • the overall length of the data packet 54 (absent the preamble 56) is variable between a minimum length of 64 bytes and a maximum length of 1518 bytes.
  • Received data packets 54 are stored in one or more buffers 70 within the packet memory 20 (see FIG. 4).
  • the size of the buffers 70 is selected such that the smallest size data packet, i.e., 64 bytes, can fit within a single buffer 70 while the largest size data packet, i.e., 1518 bytes, will require a plurality of buffers 70.
  • This buffer size selection e.g., 256 bytes per buffer in a preferred embodiment, alternatively allows the size of the packet memory 20 to be reduced or the number of packets for a given packet memory size to be increased.
  • a minimum of 219 buffers, each 256 bytes are available for each port, i.e., each MAC 18 and its associated LAN interface.
  • the packet memory size would be 1,794,048 (219 * 256 * 32), preferably implemented in a 2 Mb block of memory. Since the data packet communication rate will typically be different on each port, the buffers 70 can be allocated on the fly between ports with low data packet communication rates to those with higher demands. As each buffer 70 is allocated, data is set in a packet buffer allocation table 72 stored within the descriptor memory 22, e.g., setting one of the N bytes in the packet buffer allocation table 72 corresponding to each packet memory buffer 70. Similarly, this data is reset when a buffer 70 is deallocated, freeing the buffer 70 for future data packets.
  • pointers referred to as descriptors 74
  • receive session registers 76 which individually point to the buffers 70 that have been allocated for an incoming data packet.
  • these descriptors 74 are copied from the receive session registers 76 to one of M output queues 78 (one for each MAC port connected to the switch controller 23) that are saved in the descriptor memory 22.
  • a destination vector pointer 80 associated with a set of (1-6) descriptors 74, is stored in the output queue 78.
  • the destination vector pointer 80 points to a destination vector 82 also contained within the descriptor memory 22.
  • Each destination vector 82 contains a bit corresponding to each potential destination port, e.g., a 32 bit destination vector can be used to allocate up to 32 destination ports.
  • the destination port vectors 82 are allocated using a destination port allocation table 84 in a similar manner to the packet buffer allocation table 72.
  • a data packet 54 is directed, i.e., unicast, to a single destination address 60 on a known port, i.e., LAN.
  • a known port i.e., LAN.
  • a single bit corresponding to the one destination port is set in the destination port vector 82 and after transmission to this single port, both the buffers 70 and the destination port vector 82 are both deallocated.
  • write 86 and read 88 pointers are contained within the switch controller 23 for each of the output queues 78.
  • the queues 78 are organized as circular (ring) buffers, i.e., so that data entries following the physical end of each queue 78 are located at the physical beginning of the queue 78.
  • the write pointer 86 points to the last entry stored in the queue 78 and the read pointer 88 points to the last entry read from the queue 78.
  • the read pointer 88 does not equal the write pointer 86, it signifies that there is new data (pointing to a received data packet) within the queue 78. However, if the write pointer 86 were permitted to loop around the read pointer 88, the queue 78 would lose data.
  • the write pointer is not permitted to loop the read pointer and instead the most recently received data packet is dropped and (2) the queue size is made large enough, 256 entries long in an exemplary embodiment, to reduce the probability that a data packet would be dropped. Since the number of queue entries required for a particular data packet varies from a minimum of 2 (1 descriptor 74 plus a destination vector pointer 80) to a maximum of 7 (6 descriptors 74 plus a destination vector pointer 80), the amount of data packets 54 that can be referenced from the queue 78 also varies from 128 for a minimum size packet to 36 for a maximum size packet in an exemplary embodiment having a 256 entry queue.
  • FIG. 5 shows an exemplary block diagram of the portion of the switch controller 23 dedicated to processing received data packets.
  • This portion i.e., the receive controller 40 (preferably a state machine), is preferably comprised of multiple state machines, each of which concurrently processes portions of a receive task 90.
  • the receive controller 40 preferably a state machine
  • the receive task 90 is preferably comprised of multiple state machines, each of which concurrently processes portions of a receive task 90.
  • the receive controller 40 preferably a state machine
  • the primary state machines are (1) a main receive state machine 92 for exercising supervisory control over the other state machines, controlling the flow of data between the operational blocks, and controlling requests for packet memory buffers 70 and 64 byte subpackets according to the descriptors 74 stored in the session registers 76 (2) a receive ring state machine 94 for controlling the storage of descriptors 74 and destination vector pointers 80 into one or port queues 78, (3) a receive session state machine 96 for controlling access to the descriptors 74 stored in the session registers 76, (4) a buffer allocator 98 for allocating (and deallocating, e.g., in the case of a data packet received with a CRC error) buffers 70 in the packet memory 20, and (5) an address table state machine 100 for determining the destination port or ports according to the destination address 60 of the received data packet 54.
  • a main receive state machine 92 for exercising supervisory control over the other state machines, controlling the flow of data between the operational blocks, and controlling requests for packet memory buffers 70
  • an arbiter 102 is used to control the submission of tasks to the state machines of the receive controller 40 on a per port basis, preferably in a round robin fashion based upon status data returned from each MAC 18 and/or a data available signal 103 from the MAC receive buffer 30.
  • the arbiter 102 can submit tasks based on other criterion, e.g., the transmission speed of the port (10 Mbps, 100 Mbps, 1000 Mbps), the probability of an additional packet being received inversely based on when the last packet was received, etc.
  • the exemplary receive task 90 begins in block 104 where the task waits by periodically polling the MAC 18 for the selected port to determine if data is available.
  • a burst of data (64 bytes in this exemplary embodiment) is fetched from the selected MAC 18.
  • each MAC 18 has a receive buffer 108, e.g., a first in first out buffer (FIFO), that provides the oldest buffered data across the bidirectional MAC interface bus 31 (see FIG. 2), common to all of the MACs 18, to the MAC receive buffer 30.
  • FIFO first in first out buffer
  • the MAC receive buffer 30 is comprised of a receive first in first out buffer (FIFO) 110 under control of a receive FIFO controller 112 operating in coordination with a high speed shared data bus interface 114 to control the transfer data across the MAC interface bus 31.
  • the FIFO controller 112 independently controls this data transfer once it receives a command from the receive controller 40.
  • the MAC interface bus 31 and the data bus interface 114 are 64 bits wide. Accordingly, the data bus interface 114 is used to fetch and load the 64 byte MAC receive buffer 30 in eight 64 bit bursts.
  • the MAC receive buffer 30 can receive data at a first rate from the MACs 18 via the MAC interface bus 31 and send this received data or portion of this data toward the packet memory 20 at a second rate, i.e., the data transfers can overlap.
  • a search begins in block 118a to determine the destination port according to the data contained within the first 6 bytes of the incoming data packet, i.e., the destination address 60, using the address table state machine 100 which processes the contents of the address table 38.
  • the address table state machine 100 proceeds independently and since this task begins early in the receive task 90, sufficient time is allocated for it to complete (in block 118b where the results are stored in the receive session registers 76) before the end 68 of the data packet 54 is reached.
  • the buffer allocator 98 begins in response to receipt of the first 64 byte portion, a subpacket or subportion of a data packet, of buffer 70.
  • the associated descriptor 74 in the receive session registers 76 is updated and maintained by the concurrently operating receive session state machine 96.
  • Buffer 70 is allocated (in block 122) from the packet memory 20 whenever it is determined (e.g., in block 124) that the next subpacket will require a new 256 byte buffer, i.e., four of four 64 byte subpackets have been used.
  • the buffer allocator 98 is preferably a separate state machine that operates concurrently with the other tasks.
  • the simplified buffer allocator 98 of FIG. 6 shows buffer allocation beginning when each new buffer 70 is needed.
  • this allocation task actually begins by always having the next free buffer 70 allocated in advance so that a new buffer is always available when needed. This alternative embodiment is possible as a consequence of the use of independent state machines.
  • Each 64 byte subpacket is stored (see blocks 126 and 128) in the allocated buffer portion of the packet memory 20 described by the current session register 76 and maintained by the receive session state machine 96. Actually, this process occurs in two stages. First, the 64 byte subpacket is stored in the packet memory output buffer 36. Next, the contents of the output buffer 36 are sent to packet memory 20 via a packet memory interface 130 and the bidirectional packet memory interface bus 35 (see FIG. 2), preferably using circuitry similar to that described in reference to the MAC receive buffer 30, e.g., a controller 132 that independently controls the transfer of data into a FIFO 134 once it receives a command from the receive controller 40. In an exemplary embodiment, the bidirectional packet memory interface bus 35 is 16 bits wide.
  • the packet memory output buffer 36 is preferably implemented as a FIFO (similar to the MAC receive buffer 30), permitting data to be written into the packet memory 20 at a different rate than it is written into the packet memory output buffer 36.
  • This structure is advantageous to accommodate the different bus widths (16 bit external vs 64 bit internal) and different external and internal clock rates.
  • the descriptors 74 are updated by the receive session state machine 92 in block 136. If an end of packet (EOP) 64 has been detected in block 138, the results of the address table search of block 118 are retrieved from the address table state machine 100 in block 140 (e.g., from data stored in the receive session registers 76) and submitted to the concurrently operating ring state machine 94. In the case of a unicast message where the destination address 60 is known to the address table 38, the received data packet 54 is forwarded to the port which contains a device corresponding to the destination address 60.
  • EOP end of packet
  • the address table state machine 100 updates the address table 38 according to the contents of the source address 62 and forwards the received data packet 54 to all ports with the exception of the port which received the data packet, i.e., floods the data ports with the received data packet.
  • This task is done in block 142 by storing the descriptors 74 and the destination port vector 82 for the received data packet 54 in one or more output queues 78 in the descriptor memory 22.
  • the write pointer 86 for each of the updated output queues 78 is accordingly updated in block 144. Following updating the write pointer 86, there will be a difference between the write pointer 86 and the read pointer 88 and it is this difference that is detected by the transmit controller 42 in determining whether there are data packets to be output on each port.
  • Each of the aforementioned blocks perform portions of the receive task 90.
  • the arbiter 102 preferably allows only a portion, e.g., one task block, of the receive task 90 to complete before moving on, e.g., in a round robin manner, to complete a portion of the receive task 90 for a next port.
  • different portions of each receive task 90 can occur concurrently for each port using shared resources while maintaining data in buffers 70 within a common packet memory 20.
  • a first port could be beginning receipt of a data packet 54a into its MAC 18a
  • a second port could have received a first 64 bytes into its MAC 18b and could be transferring 64 bytes into the MAC receive buffer 30
  • a third MAC 18c could have received a second portion of a data packet into its internal FIFO 108 and could be awaiting its transfer to the MAC receive buffer 30 which is currently busy. Similar overlapping uses of the structure of the receive controller 40 occur throughout.
  • FIG. 7 shows an exemplary block diagram of the portion of the switch controller 23 dedicated to transmitting data packets.
  • This portion i.e., the transmit controller 42 (preferably a state machine), is preferably comprised of multiple state machines, each of which concurrently processes portions of a transmit task 150.
  • the transmit controller 42 preferably a state machine
  • the transmit task 150 is preferably comprised of multiple state machines, each of which concurrently processes portions of a transmit task 150.
  • different portions of the various tasks used for transmitting data to multiple MACS 18 can be performed concurrently.
  • the primary state machines are (1) a main transmit state machine 152 for exercising supervisory control over the other state machines, controlling the flow of data between the operational blocks, and controlling access to 64 byte subpackets within the buffer 70 pointed to by the current session register 76, (2) a transmit ring state machine 154 for controlling the retrieval of: descriptors 74 and the destination vector pointer 80 from port queues 78 into internal transmit session registers 156, (3) a transmit session state machine 158 for controlling access to the transmit session registers 156, and (4) a buffer deallocator 160 for deallocating buffers 70 in the packet memory 20. Additionally, an arbiter 162 is used to control the submission of tasks to the state machines of the transmit controller 42 on a per port basis, preferably in a round robin fashion.
  • the arbiter 162 can submit tasks based on other criterion, e.g., the transmission speed of the port (10 Mbps, 100 Mbps, 1000 Mbps), the probability of an additional packet being received inversely based on when the last packet was transmitted, etc.
  • other criterion e.g., the transmission speed of the port (10 Mbps, 100 Mbps, 1000 Mbps), the probability of an additional packet being received inversely based on when the last packet was transmitted, etc.
  • the exemplary transmit task 150 begins in block 164 where the task periodically (1) compares the write 86 and read 88 pointers for the selected port to determine if the selected port queue 78 has data, i.e., a data packet is available for transmission, and (2) if the MAC transmit FIFO 178 is available to receive data, e.g., by polling a space available line 165. If a data packet and the MAC transmit FIFO 178 are available, the task continues in block 166 where descriptors 74 are fetched from the port queue 78 in the descriptor memory 22 and stored in the transmit session registers 156.
  • a 64 byte subpacket is retrieved into the packet memory input buffer 34 from the buffer 70 in the packet memory 20 pointed to by the transmit session registers 156.
  • the packet memory input buffer 34 is configured as a FIFO operating in a similar manner to that described in reference to the MAC receive buffer 30, e.g., a controller 169 independently controls the transfer of data into a FIFO 170 once it receives a command from the transmit controller 42.
  • the MAC transmit buffer 32 is preferably comprised of a transmit FIFO 174 and a transmit FIFO controller 176 which independently controls the transfer of the subpacket to a transmit FIFO 178 within the designated MAC 78 via the high speed shared data bus interface 114 and the MAC interface bus 31 once it receives a command from the transmit controller 42.
  • block 180 it is determined if the current subpacket contains the EOP 68. If not, the transmit session state machine 158 is updated in block 182 to point to the next subpacket in the packet memory 20 and, in block 183, it is determined if the MAC transmit FIFO 178 is available to accept more data. If so, the process continues with block 168 fetching the next 64 byte subpacket.
  • the buffer deallocator 160 (in block 184) clears the bit corresponding to the selected port in the destination port vector 82 and the ring state machine 154 concurrently updates the read pointer 88 to deallocate the descriptors 74 in the selected port queue 78 in block 186. Providing that all bits the destination port vector 82 are now cleared, the buffer deallocator 160 (in block 188) deallocates the buffers 70 used to store the data packet. However, if bits are still set in the destination port vector 82, it signifies that other port queues 78 will still be transmitting the stored data packet and, thus, the buffers 70 are retained.
  • Each of the aforementioned blocks perform portions of the transmit task 150.
  • the arbiter 162 preferably allows only a portion, e.g., one task block, of the transmit task 150 to complete before moving on, e.g., in a round robin manner based upon status from each MAC 18 and/or status data on the space available line 165 from the MAC transmit buffer 32, to complete a portion of the transmit task 150 for a next port.
  • different portions of each transmit task 150 can occur concurrently using shared resources while operating on data packets in buffers 70 within the common packet memory 20.
  • a first MAC 18a could begin transmitting a portion of a data packet 54a from within its MAC transmit FIFO 178, a first 64 bytes of data packet 54b could have been transferred into the MAC transmit buffer 32 and could now be transferring its contents to the MAC transmit FIFO 178 in MAC 18b, and the transmit controller 42 could be awaiting the availability of the MAC transmit FIFO 178 in MAC 18c before transferring a 64 byte subpacket of the data packet 54c from the packet memory 20 to the MAC transmit buffer 32 which is currently busy. Similar overlapping uses of the structure of the transmit controller 42 occur throughout.
  • the receive 40 and transmit 42 controllers operate essentially independently, they can concurrently perform portions of their tasks for each of a plurality of LAN ports. Additionally, as previously described, the receive 40 and transmit 42 controllers are preferably comprised of a plurality of controllers, e.g., state machines, that each operate concurrently to perform portions of the receive and transmit tasks.
  • FIG. 9 shows exemplary receive 90 and transmit 150 tasks overlapping in the present invention (for simplicity, this is shown in the case where the data packet size is 64 bytes, i.e., a subpacket is the same as a packet).
  • a first receive task RX1 (on a first port) is shown placing a 64 byte packet on the MAC interface bus 31.
  • the packet is stored into the MAC receive buffer 30.
  • a buffer 70 is then allocated and pointed to by the receive session registers 76 to control the transfer of data from the packet memory 20 to the packet memory output buffer 36.
  • This sequence of operations is shown concurrently overlapping a different portion of the receive task 90 on a second port. This is possible since the receive task 90 has been subdivided to permit different portions of the task to be performed on different portions of the shared resources. Additionally, due to the buffering on the interfaces to the shared resources, each port can withstand at least a packet (64 byte) transmission delay time without any data loss.
  • multiple transmit tasks 150 can overlap each other as well as the receive tasks which can overlap each other as well as concurrent transmit tasks, e.g., TX1 data on the MAC interface bus 31 is shown overlapping TX2 data coming from the packet memory 20 and RX3 descriptor data being entered into the descriptor memory 22. Again, this is possible since each task is accessing different portions of the shared resources while maintaining intermediate data within the buffered data interfaces.
  • the described architecture of the switch controller is conducive to being implemented within a single integrated circuit package, e.g., an application specific integrated circuit (ASIC), with the packet 20 and descriptor 22 memories and the MACs 18 being external to the single integrated circuit package.
  • these external memories are commercially produced off-the-shelf RAM (or any other read/write memory) and the MACs are also readily available off the shelf devices.
  • MACS are commercially available devices that support eight ports, i.e., octal MACs. Using such a device, allows the component count to be further minimized.
  • each described aforementioned MAC 18 would then refer to one of the eight ports within such a device.

Abstract

A switch apparatus for optimizing the transfer of data packets between a plurality of local area networks (LANs). Apparatus of the present invention are comprised of multiple independent controllers, e.g., a receive controller and a transmit controller, which share common resources including a first memory (a packet memory) which stores the data packets, a second memory (a descriptor memory) which stores pointers to the stored data packets, and buffered data paths (preferably using FIFO buffers). The independent controllers operate essentially concurrently for most tasks while interleaving their use of the shared resources. Consequently, embodiments of the present invention can simultaneously receive and transmit data across multiple LAN data ports (e.g., 28 Ethernet ports comprised of 10/100 and/or 10 Mbps ports).

Description

BACKGROUND OF THE INVENTION
The present invention relates generally to a method and apparatus for transferring data packets between a plurality of local area networks (LANs) and more particularly to a switch architecture capable of high speed packet transfer between multiple data ports (e.g., 28 Ethernet ports comprised of 10/100 and/or 10 Mbps ports)
U.S. Patents directed to apparatus for switching or bridging LANs include U.S. Pat. Nos. 4,627,052; 4,715,030; 4,737,953; 4,922,503; 5,136,580, and 5,633,865.
SUMMARY OF THE INVENTION
The present invention relates to a method and apparatus for optimizing the transfer of data packets between a plurality of local area networks (LANs). Apparatus of the present invention are comprised of multiple independent controllers, e.g., a receive controller and a transmit controller, which share common resources including a first memory (a packet memory) which stores the data packets, a second memory (a descriptor memory) which stores pointers to the stored data packets, and buffered data paths (preferably using FIFO buffers). The independent controllers operate essentially concurrently for most tasks while interleaving their use of the shared resources. Consequently, embodiments of the present invention can simultaneously receive and transmit data across multiple LAN data ports (e.g., 28 Ethernet ports comprised of 10/100 and/or 10 Mbps ports).
In accordance with a preferred embodiment, a system for selectively transferring data packets between a plurality of local area networks (LANs), wherein the data packets identify destination and source addresses comprises: (1) a plurality of media access controllers (MACs) each configured for receiving and transmitting data packets to and/or from a different LAN, the MACS including at least a first MAC associated with a first LAN for receiving data packets from the first LAN and a second MAC associated with a second LAN for transmitting data packets to the second LAN, (2) a MAC receive buffer for periodically receiving at least a portion of data packets from the MACs, (3) a MAC transmit buffer capable of operating concurrently with said MAC receive buffer for periodically transmitting at least a portion of data packets directed to the MACs, (4) a shared packet memory for storing a plurality of data packets, (5) a packet memory input buffer for periodically receiving and buffering at least a portion of data packets from the packet memory, (6) a packet memory output buffer capable of operating concurrently with said packet memory input buffer for periodically buffering and transmitting at least a portion of data packets to the packet memory, (7) a shared descriptor memory for storing descriptors pointing to the data packets stored within the packet memory, (8) an address table for correlating the destination address of each data packet with one of the MACs, (9) a receive controller for causing a data packet from the first MAC to be stored in the packet memory via the MAC receive buffer and the packet memory output buffer and for generating descriptors corresponding to the locations of the data packet within the packet memory, wherein the receive controller stores the descriptors in the descriptor memory in response to the address table and the destination address of the data packet, and (10) a transmit controller capable of operating concurrently with said receive controller for causing a data packet from the packet memory to be transmitted to the second MAC via the packet memory input buffer and the MAC transmit buffer according to descriptors stored within the descriptor memory.
In accordance with a further aspect of a preferred embodiment, the system additionally includes: (1) receive session registers for storing descriptors generated by the receive controller before the descriptors are stored in the descriptor memory and (2) transmit session registers for storing descriptors retrieved from the descriptor memory, wherein the transmit controller is responsive to the retrieved descriptors within the transmit registers for selecting data packets from the packet memory for transmission to the MACs.
In accordance with a still further aspect of a preferred embodiment, the aforementioned buffers are organized as first in first out buffers (FIFOs) capable of inputting data at a first rate and outputting data at a second rate.
In accordance with still another aspect of a preferred embodiment, the aforementioned buffers and the receive and transmit controllers are all implemented within a single integrated circuit package.
The novel features of the invention are set forth with particularity in the appended claims. The invention will be best understood from the following description when read in conjunction with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 comprises a top level functional block diagram of a preferred system for optimizing the transfer of data packets between a plurality of local area networks (LANs) using a switch controller;
FIG. 2 comprises a functional block diagram of the switch controller of FIG. 1;
FIG. 3 shows the basic format of an exemplary data packet transferred by the present invention;
FIG. 4 shows a simplified block diagram of the data components allocated in the packet memory and descriptor memory of FIG. 1;
FIG. 5 shows an exemplary functional block diagram of the elements used in performing the receive task of the present invention.
FIG. 6 shows an exemplary flow chart of the receive task of the present invention which utilizes multiple concurrent state machines;
FIG. 7 shows an exemplary functional block diagram of the elements used in performing the transmit task of the present invention.
FIG. 8 shows an exemplary flow chart of the transmit task of the present invention which utilizes multiple concurrent state machines; and
FIG. 9 shows an exemplary timeline showing multiple receive and transmit tasks concurrently using different shared resources to concurrently receive and transmit data packets on a plurality of LANs.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
The present invention relates to a method and apparatus for optimizing the transfer of data packets between a plurality of local area networks (LANs).
FIG. 1 shows a functional block diagram of a preferred system 10, commonly known as a switch, for optimizing the transfer of data packets between a plurality of local area networks (LANs) e.g., 12, 14, and 16, according to data contained within each data packet (described further below in reference to FIG. 3). The system 10 is primarily comprised of (1) a plurality of media access controllers (MACs) 18 each corresponding to a different LAN 12, 14, 16 for bidirectionally transferring data packets, (2) a shared packet data memory 20 (preferably RAM and referred to as packet memory) for storing data packets prior to transferring to another LAN, (3) a shared packet descriptor memory 22 (preferably RAM and referred to as descriptor memory) for storing pointers identifying the location of the data packets in the packet memory 20, and (4) a switch controller 23 defining a receive data path 24 and a transmit data path 25 for concurrently transferring data in both directions between the MACs 18 and the packet memory 20. To accomplish this concurrent data transfer on the receive 24 and transmit 25 data paths, the switch controller 23 periodically shares resources including the packet memory 20 and the descriptor memory 22. As used herein, the term periodic is intended to refer to events occurring at regular or irregular intervals. The independent controllers within the switch controller 23 operate essentially concurrently for most tasks while interleaving their use of the shared resources. Consequently, embodiments of the present invention can simultaneously receive and transmit data across multiple LAN data ports, e.g., LAN 1 12, LAN 2 14, LAN N 16, etc., via a plurality of media access controllers 18. For example, Table I shows a simplified example of multiple receive tasks, i.e., a first data packet RX1 being received from a first LAN1 (column 2) and a second data packet RX2 being received from a second LAN2 (column 3), being processed concurrently with a third data packet TX3 being transmitted on a third LAN3 (column 4) in a time distributed manner during each time slot TX. Preferably, the time slots TX can have variable durations for each task, e.g., as needed for a particular subtask. For example, the transmit task of column 4 may overlap different portions of the receive tasks and may have unequal time slot durations.
Additionally, the media access controllers 18 preferably interface to the LANs via a PHY, i.e., a physical interface layer, (not shown). In an exemplary implementation, 28 Ethernet ports comprised of 10/100 and/or 10 Mbps ports are simultaneously supported.
              TABLE I                                                     
______________________________________                                    
T.sub.0                                                                   
    RX.sub.1 to MAC.sub.1 from                                            
                 RX.sub.2 to MAC.sub.2 from                               
                               TX.sub.3 from Packet                       
    LAN.sub.1    LAN.sub.2     Memory to MAC.sub.3 on                     
                               Transmit Data Path                         
T.sub.1                                                                   
    RX.sub.1 from MAC.sub.1 to TX.sub.3 to LAN.sub.3 from                 
    Packet Memory on           MAC.sub.3                                  
    Receive Data Path                                                     
T.sub.2          RX.sub.2 from MAC.sub.2 to                               
                 Packet Memory                                            
                 on Receive Data                                          
                 Path                                                     
______________________________________                                    
FIG. 2 shows the switch controller 23 of FIG. 1 primarily comprised of (1) a MAC receive (RCV) buffer 30 for periodically receiving and buffering at least a portion of data packets from the MACs 18 via a bidirectional MAC interface bus 31, (2) a MAC transmit (XMIT) buffer 32 capable of operating concurrently with the MAC receive buffer 30 for periodically buffering and transmitting at least a portion of data packets directed to the MACs 18 via the MAC interface bus 31, (3) a packet memory input buffer 34 for periodically receiving and buffering at least a portion of data packets from the packet memory 20 via a bidirectional packet memory interface bus 35, (4) a packet memory output buffer 36 capable of operating concurrently with the packet memory input buffer 34 for periodically buffering and transmitting at least a portion of data packets to the packet memory 20 via the packet memory interface bus 35, (5) an address table 38 for correlating the destination address of each data packet with one of the MACs 18 and its corresponding LAN, (6) a receive (RCV) controller 40 for causing a data packet from a first MAC, e.g., 18a, to be stored in the packet memory via the MAC receive buffer 30 and the packet memory output buffer 36 and for generating descriptors corresponding to the locations of each data packet within the packet memory 20, wherein the receive controller 40 stores descriptors in the descriptor memory 22 in response to the contents of the address table 38 and the destination address within the data packet, and (7) a transmit (XMIT) controller 42 for causing a data packet from the packet memory 20 to be transmitted to a second MAC, e.g., 18b, via the packet memory input buffer 34 and the MAC transmit buffer 32 according to descriptors stored within the descriptor memory 22. Additionally, embodiments of the invention preferably include a CPU interface bus 44 to a CPU 46 which can monitor and/or alter the flow of data within the switch controller 23. Associated with the CPU interface bus 44 is a CPU transmit (XMIT) buffer 48 and a CPU receive (RCV) buffer 50 which perform a similar function to that of the MAC transmit 32 and receive 30 buffers. Additionally, when the CPU interface bus 44 is present, a multiplexer 51 under control of the transmit 42 and receive 40 controllers is used to direct data between the packet memory 20 to either the MACs 18 or the CPU 46 via its associated buffered interfaces 31, 35, and 44. When not used with a CPU, the MAC receive buffer 30 is essentially coupled to the packet memory output buffer 36 via path 52 to form the receive data path 24. Similarly, the MAC transmit buffer 32 is essentially coupled to the packet memory input buffer 34 via path 53 to form the transmit data path 25.
As a consequence of the aforementioned structure, i.e., a plurality of concurrently operating buffered data paths (e.g., using buffers 30-36) sharing resources (e.g., memories 20-22), embodiments of the present invention are capable of concurrently receiving and transmitting data packets on a plurality of LANs by using a plurality of controllers (e.g., 40-42) which concurrently process portions of data packets. For example, Table II shows an expanded view of the simplified example previously shown in Table I corresponding to the preferred implementation of the receive 24 and transmit 25 data paths shown in FIG. 2.
              TABLE II                                                    
______________________________________                                    
T.sub.0                                                                   
    RX.sub.1 to MAC.sub.1 from                                            
                  RX.sub.2 to MAC.sub.2 from                              
                               TX.sub.3 Descriptors                       
    LAN.sub.1     LAN.sub.2    from Descriptor                            
                               Memory                                     
T.sub.1                                                                   
    RX.sub.1 to MAC            TX.sub.3 from Packet                       
    Receive Buffer             Memory to Packet                           
                               Memory Input                               
                               Buffer                                     
T.sub.2                                                                   
    RX.sub.1 from MAC          TX.sub.3 from Packet                       
    Receive Buffer to          Memory Input                               
    Packet Memory              Buffer to MAC                              
    Output Buffer              Transmit Buffer                            
T.sub.3                                                                   
    RX.sub.1 from Packet                                                  
                  RX.sub.2 to MAC                                         
    Memory Output Receive Buffer                                          
    Buffer to Packet                                                      
    Memory                                                                
T.sub.4                                                                   
    RX.sub.1 Descriptors to                                               
                  RX.sub.2 from MAC                                       
                               TX.sub.3 to MAC.sub.3 from                 
    Descriptor Memory                                                     
                  Receive Buffer                                          
                               MAC Transmit                               
                  to Packet    Buffer                                     
                  Memory Output                                           
                  Buffer                                                  
T.sub.5           RX.sub.2 from Packet                                    
                               TX.sub.3 from MAC.sub.3 to                 
                  Memory Output                                           
                               LAN.sub.3                                  
                  Buffer to                                               
                  Packet Memory                                           
T.sub.6           RX.sub.2 Descriptors                                    
                  to Descriptor                                           
                  Memory                                                  
______________________________________                                    
FIG. 3 shows the basic structure of an exemplary received data packet 54 (e.g., for an Ethernet data transfer protocol) comprised of (1) a preamble 56 for designating the beginning of the data packet 58 (BOP), (2) a destination address (DA) 60 for identifying a destination device present on a destination LAN, e.g., LAN 2 14, (3) a source address 62 (SA) for identifying an originating device present on a source LAN, e.g., LAN 1 12, (4) a data field 64 containing the information content of the data packet 54, and (5) a CRC field 66 at the end of the data packet (EOP 68) for performing an integrity check on the contents of the data packet 54. The overall length of the data packet 54 (absent the preamble 56) is variable between a minimum length of 64 bytes and a maximum length of 1518 bytes.
Received data packets 54 are stored in one or more buffers 70 within the packet memory 20 (see FIG. 4). Preferably the size of the buffers 70 is selected such that the smallest size data packet, i.e., 64 bytes, can fit within a single buffer 70 while the largest size data packet, i.e., 1518 bytes, will require a plurality of buffers 70. This buffer size selection, e.g., 256 bytes per buffer in a preferred embodiment, alternatively allows the size of the packet memory 20 to be reduced or the number of packets for a given packet memory size to be increased. For an exemplary implementation, a minimum of 219 buffers, each 256 bytes, are available for each port, i.e., each MAC 18 and its associated LAN interface. Accordingly for a 32 port system, the packet memory size would be 1,794,048 (219 * 256 * 32), preferably implemented in a 2 Mb block of memory. Since the data packet communication rate will typically be different on each port, the buffers 70 can be allocated on the fly between ports with low data packet communication rates to those with higher demands. As each buffer 70 is allocated, data is set in a packet buffer allocation table 72 stored within the descriptor memory 22, e.g., setting one of the N bytes in the packet buffer allocation table 72 corresponding to each packet memory buffer 70. Similarly, this data is reset when a buffer 70 is deallocated, freeing the buffer 70 for future data packets.
During the receive process, pointers, referred to as descriptors 74, are saved in receive session registers 76 which individually point to the buffers 70 that have been allocated for an incoming data packet. As described further below, once a complete data packet 54 has been successfully received, these descriptors 74 are copied from the receive session registers 76 to one of M output queues 78 (one for each MAC port connected to the switch controller 23) that are saved in the descriptor memory 22. Additionally, a destination vector pointer 80, associated with a set of (1-6) descriptors 74, is stored in the output queue 78. The destination vector pointer 80 points to a destination vector 82 also contained within the descriptor memory 22. Each destination vector 82 contains a bit corresponding to each potential destination port, e.g., a 32 bit destination vector can be used to allocate up to 32 destination ports. The destination port vectors 82 are allocated using a destination port allocation table 84 in a similar manner to the packet buffer allocation table 72.
Even in cases which require flooding, i.e., transmitting a received data packet 54 to multiple ports either because the destination port is not known or when the data packet is purposely broadcast to all ports, only a single destination port vector 82 is used for each received data packet 54. In operation, when flooding is required in a single VLAN case (i.e., when all of the physical ports of the switch are dedicated to single virtual LAN), descriptors 74 pointing to the buffers 70 which contain the received data packet 54 and the destination vector pointer 80 are stored in each of the output queues 78 (with the exception of the port associated with the LAN where the data packet originated). Similar operations occur in the case of a multicast data packet, in that descriptors 74 and a destination vector are stored in each of prescribed output queues 78. As the received data packet 54 is transmitted on each queued output port, i.e., a LAN via its associated MAC 18, the corresponding bit is cleared in the single destination port vector 82 associated with the received data packet 54. However, it is not until all of the bits in the destination port vector 82 are cleared that the commonly used buffers 70 for the stored data packet can be deallocated in the packet buffer allocation table 72. At that time, the destination port vector 82 is deallocated in the destination port allocation table 80. More commonly, a data packet 54 is directed, i.e., unicast, to a single destination address 60 on a known port, i.e., LAN. In this case, only a single bit corresponding to the one destination port is set in the destination port vector 82 and after transmission to this single port, both the buffers 70 and the destination port vector 82 are both deallocated.
Preferably, write 86 and read 88 pointers are contained within the switch controller 23 for each of the output queues 78. Preferably, the queues 78 are organized as circular (ring) buffers, i.e., so that data entries following the physical end of each queue 78 are located at the physical beginning of the queue 78. The write pointer 86 points to the last entry stored in the queue 78 and the read pointer 88 points to the last entry read from the queue 78. Whenever the read pointer 88 does not equal the write pointer 86, it signifies that there is new data (pointing to a received data packet) within the queue 78. However, if the write pointer 86 were permitted to loop around the read pointer 88, the queue 78 would lose data. Accordingly, in a preferred embodiment, (1) the write pointer is not permitted to loop the read pointer and instead the most recently received data packet is dropped and (2) the queue size is made large enough, 256 entries long in an exemplary embodiment, to reduce the probability that a data packet would be dropped. Since the number of queue entries required for a particular data packet varies from a minimum of 2 (1 descriptor 74 plus a destination vector pointer 80) to a maximum of 7 (6 descriptors 74 plus a destination vector pointer 80), the amount of data packets 54 that can be referenced from the queue 78 also varies from 128 for a minimum size packet to 36 for a maximum size packet in an exemplary embodiment having a 256 entry queue.
FIG. 5 shows an exemplary block diagram of the portion of the switch controller 23 dedicated to processing received data packets. This portion, i.e., the receive controller 40 (preferably a state machine), is preferably comprised of multiple state machines, each of which concurrently processes portions of a receive task 90. Preferably, by forming operational portions of the receive task 90 as state machines, different portions of the various tasks used for receiving data from multiple MACS 18 can be performed concurrently. The primary state machines are (1) a main receive state machine 92 for exercising supervisory control over the other state machines, controlling the flow of data between the operational blocks, and controlling requests for packet memory buffers 70 and 64 byte subpackets according to the descriptors 74 stored in the session registers 76 (2) a receive ring state machine 94 for controlling the storage of descriptors 74 and destination vector pointers 80 into one or port queues 78, (3) a receive session state machine 96 for controlling access to the descriptors 74 stored in the session registers 76, (4) a buffer allocator 98 for allocating (and deallocating, e.g., in the case of a data packet received with a CRC error) buffers 70 in the packet memory 20, and (5) an address table state machine 100 for determining the destination port or ports according to the destination address 60 of the received data packet 54. Additionally, an arbiter 102 is used to control the submission of tasks to the state machines of the receive controller 40 on a per port basis, preferably in a round robin fashion based upon status data returned from each MAC 18 and/or a data available signal 103 from the MAC receive buffer 30. Alternatively, the arbiter 102 can submit tasks based on other criterion, e.g., the transmission speed of the port (10 Mbps, 100 Mbps, 1000 Mbps), the probability of an additional packet being received inversely based on when the last packet was received, etc.
As shown in FIG. 6, the exemplary receive task 90 (shown for a single port as selected by the arbiter 102 once data is first detected in the selected MAC 18) begins in block 104 where the task waits by periodically polling the MAC 18 for the selected port to determine if data is available. In block 106, a burst of data (64 bytes in this exemplary embodiment) is fetched from the selected MAC 18. Preferably, each MAC 18 has a receive buffer 108, e.g., a first in first out buffer (FIFO), that provides the oldest buffered data across the bidirectional MAC interface bus 31 (see FIG. 2), common to all of the MACs 18, to the MAC receive buffer 30. Preferably, the MAC receive buffer 30 is comprised of a receive first in first out buffer (FIFO) 110 under control of a receive FIFO controller 112 operating in coordination with a high speed shared data bus interface 114 to control the transfer data across the MAC interface bus 31. In a preferred embodiment, the FIFO controller 112 independently controls this data transfer once it receives a command from the receive controller 40. In this exemplary embodiment, the MAC interface bus 31 and the data bus interface 114 are 64 bits wide. Accordingly, the data bus interface 114 is used to fetch and load the 64 byte MAC receive buffer 30 in eight 64 bit bursts. Since the MAC receive buffer 30 is preferably organized as a FIFO, it can receive data at a first rate from the MACs 18 via the MAC interface bus 31 and send this received data or portion of this data toward the packet memory 20 at a second rate, i.e., the data transfers can overlap.
Once the first 64 bytes have been received into the MAC receive buffer 108 in block 116, a search begins in block 118a to determine the destination port according to the data contained within the first 6 bytes of the incoming data packet, i.e., the destination address 60, using the address table state machine 100 which processes the contents of the address table 38. The address table state machine 100 proceeds independently and since this task begins early in the receive task 90, sufficient time is allocated for it to complete (in block 118b where the results are stored in the receive session registers 76) before the end 68 of the data packet 54 is reached.
It is then determined in block 120 whether the packet memory output buffer 36 is available. In the exemplary receive task 90 of FIG. 6, it is shown that the buffer allocator 98 begins in response to receipt of the first 64 byte portion, a subpacket or subportion of a data packet, of buffer 70. Once buffer 70 is allocated by the buffer allocator 98, the associated descriptor 74 in the receive session registers 76 is updated and maintained by the concurrently operating receive session state machine 96. There are up to four 64 byte subpackets per 256 byte buffer 70 and this is tracked within a subpacket field of each descriptor 74 in the receive session registers 76. Buffer 70 is allocated (in block 122) from the packet memory 20 whenever it is determined (e.g., in block 124) that the next subpacket will require a new 256 byte buffer, i.e., four of four 64 byte subpackets have been used. Again, the buffer allocator 98 is preferably a separate state machine that operates concurrently with the other tasks. The simplified buffer allocator 98 of FIG. 6 shows buffer allocation beginning when each new buffer 70 is needed. However, in an alternative preferred embodiment, this allocation task actually begins by always having the next free buffer 70 allocated in advance so that a new buffer is always available when needed. This alternative embodiment is possible as a consequence of the use of independent state machines.
Each 64 byte subpacket is stored (see blocks 126 and 128) in the allocated buffer portion of the packet memory 20 described by the current session register 76 and maintained by the receive session state machine 96. Actually, this process occurs in two stages. First, the 64 byte subpacket is stored in the packet memory output buffer 36. Next, the contents of the output buffer 36 are sent to packet memory 20 via a packet memory interface 130 and the bidirectional packet memory interface bus 35 (see FIG. 2), preferably using circuitry similar to that described in reference to the MAC receive buffer 30, e.g., a controller 132 that independently controls the transfer of data into a FIFO 134 once it receives a command from the receive controller 40. In an exemplary embodiment, the bidirectional packet memory interface bus 35 is 16 bits wide. Accordingly, the transfer of the 64 byte subpacket requires a burst transfer of thirty-two 16 bit quantities. The packet memory output buffer 36 is preferably implemented as a FIFO (similar to the MAC receive buffer 30), permitting data to be written into the packet memory 20 at a different rate than it is written into the packet memory output buffer 36. This structure is advantageous to accommodate the different bus widths (16 bit external vs 64 bit internal) and different external and internal clock rates.
Once the process of sending each subpacket to the packet memory commences (in blocks 126 and 128), the descriptors 74 are updated by the receive session state machine 92 in block 136. If an end of packet (EOP) 64 has been detected in block 138, the results of the address table search of block 118 are retrieved from the address table state machine 100 in block 140 (e.g., from data stored in the receive session registers 76) and submitted to the concurrently operating ring state machine 94. In the case of a unicast message where the destination address 60 is known to the address table 38, the received data packet 54 is forwarded to the port which contains a device corresponding to the destination address 60. Otherwise, the address table state machine 100 updates the address table 38 according to the contents of the source address 62 and forwards the received data packet 54 to all ports with the exception of the port which received the data packet, i.e., floods the data ports with the received data packet. This task is done in block 142 by storing the descriptors 74 and the destination port vector 82 for the received data packet 54 in one or more output queues 78 in the descriptor memory 22. The write pointer 86 for each of the updated output queues 78 is accordingly updated in block 144. Following updating the write pointer 86, there will be a difference between the write pointer 86 and the read pointer 88 and it is this difference that is detected by the transmit controller 42 in determining whether there are data packets to be output on each port.
Each of the aforementioned blocks perform portions of the receive task 90. The arbiter 102 preferably allows only a portion, e.g., one task block, of the receive task 90 to complete before moving on, e.g., in a round robin manner, to complete a portion of the receive task 90 for a next port. As a consequence of the aforementioned structure, different portions of each receive task 90 can occur concurrently for each port using shared resources while maintaining data in buffers 70 within a common packet memory 20. For example, a first port could be beginning receipt of a data packet 54a into its MAC 18a, a second port could have received a first 64 bytes into its MAC 18b and could be transferring 64 bytes into the MAC receive buffer 30, and a third MAC 18c could have received a second portion of a data packet into its internal FIFO 108 and could be awaiting its transfer to the MAC receive buffer 30 which is currently busy. Similar overlapping uses of the structure of the receive controller 40 occur throughout.
FIG. 7 shows an exemplary block diagram of the portion of the switch controller 23 dedicated to transmitting data packets. This portion, i.e., the transmit controller 42 (preferably a state machine), is preferably comprised of multiple state machines, each of which concurrently processes portions of a transmit task 150. Preferably, by forming operational portions of the transmit task 150 as state machines, different portions of the various tasks used for transmitting data to multiple MACS 18 can be performed concurrently. The primary state machines are (1) a main transmit state machine 152 for exercising supervisory control over the other state machines, controlling the flow of data between the operational blocks, and controlling access to 64 byte subpackets within the buffer 70 pointed to by the current session register 76, (2) a transmit ring state machine 154 for controlling the retrieval of: descriptors 74 and the destination vector pointer 80 from port queues 78 into internal transmit session registers 156, (3) a transmit session state machine 158 for controlling access to the transmit session registers 156, and (4) a buffer deallocator 160 for deallocating buffers 70 in the packet memory 20. Additionally, an arbiter 162 is used to control the submission of tasks to the state machines of the transmit controller 42 on a per port basis, preferably in a round robin fashion. Alternatively, the arbiter 162 can submit tasks based on other criterion, e.g., the transmission speed of the port (10 Mbps, 100 Mbps, 1000 Mbps), the probability of an additional packet being received inversely based on when the last packet was transmitted, etc.
As shown in FIG. 8, the exemplary transmit task 150 (shown for a single port as selected by the arbiter 162 once data is first detected in the selected port queue 78) begins in block 164 where the task periodically (1) compares the write 86 and read 88 pointers for the selected port to determine if the selected port queue 78 has data, i.e., a data packet is available for transmission, and (2) if the MAC transmit FIFO 178 is available to receive data, e.g., by polling a space available line 165. If a data packet and the MAC transmit FIFO 178 are available, the task continues in block 166 where descriptors 74 are fetched from the port queue 78 in the descriptor memory 22 and stored in the transmit session registers 156. Next, in block 168, a 64 byte subpacket is retrieved into the packet memory input buffer 34 from the buffer 70 in the packet memory 20 pointed to by the transmit session registers 156. Preferably the packet memory input buffer 34 is configured as a FIFO operating in a similar manner to that described in reference to the MAC receive buffer 30, e.g., a controller 169 independently controls the transfer of data into a FIFO 170 once it receives a command from the transmit controller 42. In block 171, it is determined if the MAC transmit buffer 32 is available and, if so, a 64 byte subpacket is transferred to the MAC transmit buffer 32. The MAC transmit buffer 32 is preferably comprised of a transmit FIFO 174 and a transmit FIFO controller 176 which independently controls the transfer of the subpacket to a transmit FIFO 178 within the designated MAC 78 via the high speed shared data bus interface 114 and the MAC interface bus 31 once it receives a command from the transmit controller 42. In block 180, it is determined if the current subpacket contains the EOP 68. If not, the transmit session state machine 158 is updated in block 182 to point to the next subpacket in the packet memory 20 and, in block 183, it is determined if the MAC transmit FIFO 178 is available to accept more data. If so, the process continues with block 168 fetching the next 64 byte subpacket. If the EOP 68 has been detected, the buffer deallocator 160 (in block 184) clears the bit corresponding to the selected port in the destination port vector 82 and the ring state machine 154 concurrently updates the read pointer 88 to deallocate the descriptors 74 in the selected port queue 78 in block 186. Providing that all bits the destination port vector 82 are now cleared, the buffer deallocator 160 (in block 188) deallocates the buffers 70 used to store the data packet. However, if bits are still set in the destination port vector 82, it signifies that other port queues 78 will still be transmitting the stored data packet and, thus, the buffers 70 are retained.
Each of the aforementioned blocks perform portions of the transmit task 150. The arbiter 162 preferably allows only a portion, e.g., one task block, of the transmit task 150 to complete before moving on, e.g., in a round robin manner based upon status from each MAC 18 and/or status data on the space available line 165 from the MAC transmit buffer 32, to complete a portion of the transmit task 150 for a next port. As a consequence of the aforementioned structure, different portions of each transmit task 150 can occur concurrently using shared resources while operating on data packets in buffers 70 within the common packet memory 20. For example, a first MAC 18a could begin transmitting a portion of a data packet 54a from within its MAC transmit FIFO 178, a first 64 bytes of data packet 54b could have been transferred into the MAC transmit buffer 32 and could now be transferring its contents to the MAC transmit FIFO 178 in MAC 18b, and the transmit controller 42 could be awaiting the availability of the MAC transmit FIFO 178 in MAC 18c before transferring a 64 byte subpacket of the data packet 54c from the packet memory 20 to the MAC transmit buffer 32 which is currently busy. Similar overlapping uses of the structure of the transmit controller 42 occur throughout.
Since the receive 40 and transmit 42 controllers operate essentially independently, they can concurrently perform portions of their tasks for each of a plurality of LAN ports. Additionally, as previously described, the receive 40 and transmit 42 controllers are preferably comprised of a plurality of controllers, e.g., state machines, that each operate concurrently to perform portions of the receive and transmit tasks. For example, FIG. 9 shows exemplary receive 90 and transmit 150 tasks overlapping in the present invention (for simplicity, this is shown in the case where the data packet size is 64 bytes, i.e., a subpacket is the same as a packet). A first receive task RX1 (on a first port) is shown placing a 64 byte packet on the MAC interface bus 31. Next, the packet is stored into the MAC receive buffer 30. A buffer 70 is then allocated and pointed to by the receive session registers 76 to control the transfer of data from the packet memory 20 to the packet memory output buffer 36. This sequence of operations is shown concurrently overlapping a different portion of the receive task 90 on a second port. This is possible since the receive task 90 has been subdivided to permit different portions of the task to be performed on different portions of the shared resources. Additionally, due to the buffering on the interfaces to the shared resources, each port can withstand at least a packet (64 byte) transmission delay time without any data loss. Similarly, multiple transmit tasks 150, e.g., TX1, TX2, etc., can overlap each other as well as the receive tasks which can overlap each other as well as concurrent transmit tasks, e.g., TX1 data on the MAC interface bus 31 is shown overlapping TX2 data coming from the packet memory 20 and RX3 descriptor data being entered into the descriptor memory 22. Again, this is possible since each task is accessing different portions of the shared resources while maintaining intermediate data within the buffered data interfaces.
The described architecture of the switch controller (see FIG. 2) is conducive to being implemented within a single integrated circuit package, e.g., an application specific integrated circuit (ASIC), with the packet 20 and descriptor 22 memories and the MACs 18 being external to the single integrated circuit package. In a preferred embodiment, these external memories are commercially produced off-the-shelf RAM (or any other read/write memory) and the MACs are also readily available off the shelf devices. For example, currently used MACS are commercially available devices that support eight ports, i.e., octal MACs. Using such a device, allows the component count to be further minimized. However, it should be noted that for descriptive purposes each described aforementioned MAC 18 would then refer to one of the eight ports within such a device.
Although the present invention has been described in detail with reference only to the presently-preferred embodiments, those of ordinary skill in the art will appreciate that various modifications can be made without departing from the invention. For example, while Ethernet type LANs have been described, one of ordinary skill in the art will appreciate that the present invention is applicable to other data transfer protocols as well. Accordingly, the invention is defined by the following claims.

Claims (12)

We claim:
1. A system for selectively transferring data packets between a plurality of local area networks (LANs), wherein said data packets identify destination and source addresses, said system comprising:
a plurality of media access controllers (MACs) each configured for receiving and transmitting data packets to and/or from a different LAN; said MACS including at least a first MAC associated with a first LAN for receiving data packets from said first LAN and a second MAC associated with a second LAN for transmitting data packets to said second LAN;
a shared data bus;
each of said MACs connected to said shared data bus for presenting data packets for transfer across said data bus;
a MAC receive buffer including a controller for transferring data packets presented by said MACs across shared data bus and for buffering at least a portion of said transferred data packets;
a MAC transmit buffer including a controller capable of operating concurrently with said MAC receive buffer controller for transferring data packets across said shared data bus and for buffering at least a portion of said transferred data packets;
a shared packet memory for storing a plurality of data packets;
a packet memory input buffer for periodically receiving and buffering at least a portion of data packets from said packet memory;
a packet memory output buffer capable of operating concurrently with said packet memory input buffer for periodically buffering and transmitting at least a portion of data packets to said packet memory;
a shared descriptor memory for storing descriptors pointing to said data packets stored within said packet memory;
an address table for correlating the destination address of each data packet with one of said MACs;
a receive controller for transferring a data packet from said MAC receive buffer to said packet memory via said packet memory output buffer and for generating descriptors corresponding to the locations of said data packet within said packet memory; wherein said receive controller stores said descriptors in said descriptor memory in response to said address table and the destination address of said data packet; and
a transmit controller capable of operating concurrently with said receive controller for transferring a data packet from said packet memory to said MAC transmit buffer via said packet memory input buffer according to descriptors stored within said descriptor memory.
2. The system of claim 1 wherein each of said receive and transmit controllers are state machines.
3. The system of claim 1 wherein said MAC receive buffer, said MAC transmit buffer, said packet memory input buffer, said packet memory output buffer, said receive controller, and said transmit controller are implemented within a single integrated circuit package.
4. The system of claim 1 additionally comprising:
receive session registers for storing descriptors generated by said receive controller before said descriptors are stored in said descriptor memory;
transmit session registers for storing descriptors retrieved from said descriptor memory; and wherein
said transmit controller is responsive to said retrieved descriptors within said transmit session registers for selecting data packets from said packet memory for transmission to said MACs.
5. The system of claim 4 wherein said MAC receive buffer, said MAC transmit buffer, said packet memory input buffer, said packet memory output buffer, said receive controller, said transmit controller, said receive session registers, and said transmit session registers are implemented within a single integrated circuit package.
6. The system of claim 1 wherein said MAC receive buffer, said MAC transmit buffer, said packet memory input buffer, said packet memory output buffer are each organized as first in first out buffers (FIFOs) capable of inputting data at a first rate and outputting data at a second rate.
7. In a system comprised of (1) a plurality of media access controllers (MACs) each configured for receiving and transmitting data packets to and/or from a different local area networks (LAN), said MACS including at least a first MAC associated with a first LAN for receiving and buffering data packets from said first LAN and a second MAC associated with a second LAN for transmitting data packets to said second LAN, (2) a shared packet memory for storing a plurality of data packets, and (3) a shared descriptor memory for storing descriptors pointing to said data packets stored within said packet memory, wherein said data packets identify destination and source addresses, a switch controller for selectively transferring data packets between a plurality of LANs via said MACS, said switch controller comprising:
a shared data bus;
each of said MACs connected to said shared data bus for presenting data packets for transfer across said data bus;
a MAC receive buffer including a controller for transferring data packets presented by said MACs across shared data bus and for buffering at least a portion of said transferred data packets;
a MAC transmit buffer including a controller capable of operating concurrently with said MAC receive buffer controller for transferring data packets across said shared data bus and for buffering at least a portion of said transferred data packets;
a packet memory input buffer for periodically receiving and buffering at least a portion of data packets from said packet memory;
a packet memory output buffer capable of operating concurrently with said packet memory input buffer for periodically buffering and transmitting at least a portion of data packets to said packet memory;
an address table for correlating the destination address of each data packet with one of said MACs;
a receive controller for transferring a data packet from said first MAC to be stored in said packet memory via said MAC receive buffer to said packet memory via said packet memory output buffer and for generating descriptors corresponding to the locations of said data packet within said packet memory; wherein said receive controller stores said descriptors in said descriptor memory in response to said address table and the destination address of said data packet; and
a transmit controller capable of operating concurrently with said receive controller for transferring a data packet from said packet memory to said MAC transmit buffer via said packet memory input buffer according to descriptors stored within said descriptor memory.
8. The switch controller of claim 7 wherein each of said receive and transmit controllers are state machines.
9. The switch controller of claim 7 wherein said MAC receive buffer, said MAC transmit buffer, said packet memory input buffer, said packet memory output buffer, said receive controller, and said transmit controller are implemented within a single integrated circuit package.
10. The switch controller of claim 7 additionally comprising:
receive session registers for storing descriptors generated by said receive controller before said descriptors are stored in said descriptor memory;
transmit session registers for storing descriptors retrieved from said descriptor memory; and wherein
said transmit controller is responsive to said retrieved descriptors within said transmit session registers for selecting data packets from said packet memory for transmission to said MACs.
11. The switch controller of claim 10 wherein said MAC receive buffer, said MAC transmit buffer, said packet memory input buffer, said packet memory output buffer, said receive controller, said transmit controller, said receive session registers, and said transmit session registers are implemented within a single integrated circuit package.
12. The switch controller of claim 7 wherein said MAC receive buffer, said MAC transmit buffer, said packet memory input buffer, said packet memory output buffer are each organized as first in first out buffers (FIFOs) capable of inputting data at a first rate and outputting data at a second rate.
US09/096,110 1998-06-11 1998-06-11 Method and apparatus for optimizing the transfer of data packets between local area networks Expired - Lifetime US6067300A (en)

Priority Applications (6)

Application Number Priority Date Filing Date Title
US09/096,110 US6067300A (en) 1998-06-11 1998-06-11 Method and apparatus for optimizing the transfer of data packets between local area networks
EP99927458A EP1086552B1 (en) 1998-06-11 1999-06-10 Optimizing the transfer of data packets between lans
CA002328220A CA2328220C (en) 1998-06-11 1999-06-10 Optimizing the transfer of data packets between lans
PCT/US1999/013166 WO1999065193A1 (en) 1998-06-11 1999-06-10 Optimizing the transfer of data packets between lans
AU44356/99A AU767085B2 (en) 1998-06-11 1999-06-10 Optimizing the transfer of data packets between LANs
DE69941697T DE69941697D1 (en) 1998-06-11 1999-06-10 TRANSFER OPTIMIZATION OF DATA PACKAGES BETWEEN LANS

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US09/096,110 US6067300A (en) 1998-06-11 1998-06-11 Method and apparatus for optimizing the transfer of data packets between local area networks

Publications (1)

Publication Number Publication Date
US6067300A true US6067300A (en) 2000-05-23

Family

ID=22255404

Family Applications (1)

Application Number Title Priority Date Filing Date
US09/096,110 Expired - Lifetime US6067300A (en) 1998-06-11 1998-06-11 Method and apparatus for optimizing the transfer of data packets between local area networks

Country Status (6)

Country Link
US (1) US6067300A (en)
EP (1) EP1086552B1 (en)
AU (1) AU767085B2 (en)
CA (1) CA2328220C (en)
DE (1) DE69941697D1 (en)
WO (1) WO1999065193A1 (en)

Cited By (88)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6252879B1 (en) 1997-09-17 2001-06-26 Sony Corporation Single counter for controlling multiple finite state machines in a multi-port bridge for local area network
US6301256B1 (en) 1997-09-17 2001-10-09 Sony Corporation Selection technique for preventing a source port from becoming a destination port in a multi-port bridge for a local area network
US6308218B1 (en) 1997-09-17 2001-10-23 Sony Corporation Address look-up mechanism in a multi-port bridge for a local area network
US6341313B1 (en) * 1998-02-02 2002-01-22 Nec Corporation Flow controlling method and apparatus for network between processors
WO2002015496A1 (en) * 2000-08-11 2002-02-21 Paion Company, Limited Novel switch fabric chipset system and method
US6363067B1 (en) 1997-09-17 2002-03-26 Sony Corporation Staged partitioned communication bus for a multi-port bridge for a local area network
US6401146B1 (en) * 1998-02-06 2002-06-04 Lg Semicon Co., Ltd. Device and method for controlling PCI ethernet
US6442168B1 (en) 1997-09-17 2002-08-27 Sony Corporation High speed bus structure in a multi-port bridge for a local area network
US6446173B1 (en) 1997-09-17 2002-09-03 Sony Corporation Memory controller in a multi-port bridge for a local area network
US20020172195A1 (en) * 2001-03-23 2002-11-21 Pekkala Richard E. Apparatus amd method for disparate fabric data and transaction buffering within infiniband device
US20030067934A1 (en) * 2001-09-28 2003-04-10 Hooper Donald F. Multiprotocol decapsulation/encapsulation control structure and packet protocol conversion method
US6577625B1 (en) * 1998-06-22 2003-06-10 Winbond Electronics Corp. Ethernet switch with a share memory structure and method for sharing memory
US20030110271A1 (en) * 2001-08-31 2003-06-12 Ramkumar Jayam Systems and methods for high speed data transmission using TCP/IP
US20030115338A1 (en) * 2001-08-31 2003-06-19 Ramkumar Jayam Apparatus and methods for receiving data at high speed using TCP/IP
US6584517B1 (en) * 1999-07-02 2003-06-24 Cypress Semiconductor Corp. Circuit and method for supporting multicast/broadcast operations in multi-queue storage devices
US20030120789A1 (en) * 2001-10-22 2003-06-26 Neil Hepworth Real time control protocol session matching
US6601150B1 (en) * 1999-05-10 2003-07-29 Nortel Networks Limited Memory management technique for maintaining packet order in a packet processing system
US20030145173A1 (en) * 2002-01-25 2003-07-31 Wilkinson Hugh M. Context pipelines
US20030156597A1 (en) * 2002-02-21 2003-08-21 Sun Microsystems, Inc. Method and apparatus for speculative arbitration
US6617879B1 (en) 1997-09-17 2003-09-09 Sony Corporation Transparently partitioned communication bus for multi-port bridge for a local area network
US6654811B1 (en) * 2000-04-13 2003-11-25 Nokia Inc. Backpressure arrangement in client-server environment
US20030223431A1 (en) * 2002-04-11 2003-12-04 Chavez David L. Emergency bandwidth allocation with an RSVP-like protocol
US6701393B1 (en) * 2002-06-27 2004-03-02 Emc Corporation Systems and methods for managing storage location descriptors
US20040073690A1 (en) * 2002-09-30 2004-04-15 Neil Hepworth Voice over IP endpoint call admission
US20040073641A1 (en) * 2002-09-30 2004-04-15 Muneyb Minhazuddin Instantaneous user initiation voice quality feedback
US20040073692A1 (en) * 2002-09-30 2004-04-15 Gentle Christopher R. Packet prioritization and associated bandwidth and buffer management techniques for audio over IP
US20040073778A1 (en) * 1999-08-31 2004-04-15 Adiletta Matthew J. Parallel processor architecture
US6724759B1 (en) 2000-08-11 2004-04-20 Paion Company, Limited System, method and article of manufacture for transferring a packet from a port controller to a switch fabric in a switch fabric chipset system
US6731631B1 (en) 2000-08-11 2004-05-04 Paion Company, Limited System, method and article of manufacture for updating a switching table in a switch fabric chipset system
US6738384B1 (en) 1997-09-17 2004-05-18 Sony Corporation Technique for optimizing cut-through for broadcast and multi-cast packets in a multi-port bridge for a local area network
US6748445B1 (en) * 2000-02-01 2004-06-08 Microsoft Corporation System and method for exchanging data
US20040148555A1 (en) * 2003-01-24 2004-07-29 Dennis Blackburn Apparatus and method for accommodating loss of signal
US6775283B1 (en) * 1999-11-16 2004-08-10 Advanced Micro Devices, Inc. Passing vlan information through descriptors
US6781986B1 (en) * 1999-06-25 2004-08-24 Nortel Networks Limited Scalable high capacity switch architecture method, apparatus and system
US6804731B1 (en) 2000-08-11 2004-10-12 Paion Company, Limited System, method and article of manufacture for storing an incoming datagram in switch matrix in a switch fabric chipset system
US20050063407A1 (en) * 2003-09-23 2005-03-24 Samsung Electronics Co., Ltd. Apparatus and method for maintaining high-speed forwarding tables in a massively parallel router
US20050138212A1 (en) * 2003-12-23 2005-06-23 Intel Corporation Page-aware descriptor management
US20050144338A1 (en) * 2003-02-27 2005-06-30 Fujitsu Limited Data transfer apparatus
US20050163137A1 (en) * 2004-01-28 2005-07-28 Wakumoto Shaun K. Switching mesh with broadcast path redundancy
US20050216710A1 (en) * 2002-01-17 2005-09-29 Wilkinson Hugh M Iii Parallel processor with functional pipeline providing programming engines by supporting multiple contexts and critical section
US20050216616A1 (en) * 2004-03-23 2005-09-29 Intel Corporation Inbound packet placement in host memory
US20050216822A1 (en) * 2002-02-18 2005-09-29 Hiroshi Kyusojin Data communication system, data communication management device and method, and computer program
US6952824B1 (en) 1999-12-30 2005-10-04 Intel Corporation Multi-threaded sequenced receive for fast network port stream of packets
US20050226147A1 (en) * 2004-04-06 2005-10-13 Oki Electric Industry Co., Ltd. Relay device
US6975626B1 (en) 2000-03-31 2005-12-13 Sun Microsystems, Inc. Switched network for low latency communication
US7020161B1 (en) 2000-03-31 2006-03-28 Sun Microsystems, Inc. Prescheduling arbitrated resources
US20060077995A1 (en) * 1999-01-27 2006-04-13 Broadcom Corporation Apparatus for ethernet PHY/MAC communication
US20060095611A1 (en) * 2004-11-02 2006-05-04 Standard Microsystems Corporation Hardware supported peripheral component memory alignment method
US7061929B1 (en) * 2000-03-31 2006-06-13 Sun Microsystems, Inc. Data network with independent transmission channels
US7065580B1 (en) 2000-03-31 2006-06-20 Sun Microsystems, Inc. Method and apparatus for a pipelined network
US20060230119A1 (en) * 2005-04-08 2006-10-12 Neteffect, Inc. Apparatus and method for packet transmission over a high speed network supporting remote direct memory access operations
US7158520B1 (en) 2002-03-22 2007-01-02 Juniper Networks, Inc. Mailbox registers for synchronizing header processing execution
US7180893B1 (en) 2002-03-22 2007-02-20 Juniper Networks, Inc. Parallel layer 2 and layer 3 processing components in a network router
US20070086480A1 (en) * 1999-07-30 2007-04-19 Uri Elzur Associating a packet with a flow
US7212530B1 (en) 2002-03-22 2007-05-01 Juniper Networks, Inc. Optimized buffer loading for packet header processing
US7215662B1 (en) 2002-03-22 2007-05-08 Juniper Networks, Inc. Logical separation and accessing of descriptor memories
US7236501B1 (en) 2002-03-22 2007-06-26 Juniper Networks, Inc. Systems and methods for handling packet fragmentation
US7239630B1 (en) * 2002-03-22 2007-07-03 Juniper Networks, Inc. Dedicated processing resources for packet header generation
US20070165672A1 (en) * 2006-01-19 2007-07-19 Neteffect, Inc. Apparatus and method for stateless CRC calculation
US20070198720A1 (en) * 2006-02-17 2007-08-23 Neteffect, Inc. Method and apparatus for a interfacing device drivers to a single multi-function adapter
US20070208820A1 (en) * 2006-02-17 2007-09-06 Neteffect, Inc. Apparatus and method for out-of-order placement and in-order completion reporting of remote direct memory access operations
US20070226386A1 (en) * 2006-02-17 2007-09-27 Neteffect, Inc. Method and apparatus for using a single multi-function adapter with different operating systems
US20070226750A1 (en) * 2006-02-17 2007-09-27 Neteffect, Inc. Pipelined processing of RDMA-type network transactions
US7283528B1 (en) 2002-03-22 2007-10-16 Raymond Marcelino Manese Lim On the fly header checksum processing using dedicated logic
US20080043750A1 (en) * 2006-01-19 2008-02-21 Neteffect, Inc. Apparatus and method for in-line insertion and removal of markers
US7480706B1 (en) * 1999-12-30 2009-01-20 Intel Corporation Multi-threaded round-robin receive for fast network port
US7617337B1 (en) 2007-02-06 2009-11-10 Avaya Inc. VoIP quality tradeoff system
US20100002717A1 (en) * 2002-05-03 2010-01-07 Conexant, Inc. Partitioned Medium Access Control Implementation
US7751402B2 (en) 1999-12-29 2010-07-06 Intel Corporation Method and apparatus for gigabit packet assignment for multithreaded packet processing
USRE41849E1 (en) 1999-12-22 2010-10-19 Intel Corporation Parallel multi-threaded processing
USRE41919E1 (en) 2003-06-25 2010-11-09 Steve Olivier Rapid decryption of data by key synchronization and indexing
US7899052B1 (en) * 1999-01-27 2011-03-01 Broadcom Corporation Memory structure for resolving addresses in a packet-based network switch
US7978827B1 (en) 2004-06-30 2011-07-12 Avaya Inc. Automatic configuration of call handling based on end-user needs and characteristics
US20120033673A1 (en) * 2010-08-06 2012-02-09 Deepak Goel Systems and methods for a para-vitualized driver in a multi-core virtual packet engine device
US20120137102A1 (en) * 2010-11-30 2012-05-31 Ramkumar Perumanam Consumer approach based memory buffer optimization for multimedia applications
US20120143979A1 (en) * 2005-03-24 2012-06-07 Olderdissen Jan Ra Protocol stack using shared memory
US8218751B2 (en) 2008-09-29 2012-07-10 Avaya Inc. Method and apparatus for identifying and eliminating the source of background noise in multi-party teleconferences
US8316191B2 (en) 1999-08-31 2012-11-20 Intel Corporation Memory controllers for processor having multiple programmable units
US8396063B1 (en) * 2003-05-30 2013-03-12 Redpine Signals, Inc. Multi-thread media access controller with per-thread MAC address
US20130128739A1 (en) * 2011-11-21 2013-05-23 Broadcom Corporation Method and Device to Control Communication with Multiple Networks Based on Respective Quality of Service Requirements
US20130198434A1 (en) * 2012-01-26 2013-08-01 Nokia Corporation Apparatus and Method to Provide Cache Move With Non-Volatile Mass Memory System
US20130290471A1 (en) * 2012-04-27 2013-10-31 Rajat Venkatesh Managing transfer of data from a source to a destination machine cluster
US8738886B2 (en) 1999-12-27 2014-05-27 Intel Corporation Memory mapping in a processor having multiple programmable units
US20150163141A1 (en) * 2013-12-09 2015-06-11 International Business Machines Corporation FIFO Affinity for Multiple Ports
US9164804B2 (en) 2012-06-20 2015-10-20 Memory Technologies Llc Virtual memory module
US9208078B2 (en) 2009-06-04 2015-12-08 Memory Technologies Llc Apparatus and method to share host system RAM with mass storage memory RAM
US9311226B2 (en) 2012-04-20 2016-04-12 Memory Technologies Llc Managing operational state data of a memory module using host memory in association with state change
US9367486B2 (en) 2008-02-28 2016-06-14 Memory Technologies Llc Extended utilization area for a memory device

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE10119754B4 (en) 2001-04-23 2007-11-29 Infineon Technologies Ag Method and device for storing data packets
US9116820B2 (en) 2012-08-28 2015-08-25 Memory Technologies Llc Dynamic central cache memory

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5434976A (en) * 1992-09-28 1995-07-18 Standard Microsystems Corporation Communications controller utilizing an external buffer memory with plural channels between a host and network interface operating independently for transferring packets between protocol layers
US5448565A (en) * 1992-11-12 1995-09-05 International Business Machines Corp. Multiport LAN bridge
US5633865A (en) * 1995-03-31 1997-05-27 Netvantage Apparatus for selectively transferring data packets between local area networks
US5758075A (en) * 1994-07-29 1998-05-26 International Business Machines Corporation Multimedia communication apparatus and methods
US5764895A (en) * 1995-01-11 1998-06-09 Sony Corporation Method and apparatus for directing data packets in a local area network device having a plurality of ports interconnected by a high-speed communication bus
US5805816A (en) * 1992-05-12 1998-09-08 Compaq Computer Corp. Network packet switch using shared memory for repeating and bridging packets at media rate
US5878028A (en) * 1996-06-06 1999-03-02 Advanced Micro Devices, Inc. Data structure to support multiple transmit packets for high performance

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5058110A (en) * 1989-05-03 1991-10-15 Ultra Network Technologies Protocol processor

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5805816A (en) * 1992-05-12 1998-09-08 Compaq Computer Corp. Network packet switch using shared memory for repeating and bridging packets at media rate
US5434976A (en) * 1992-09-28 1995-07-18 Standard Microsystems Corporation Communications controller utilizing an external buffer memory with plural channels between a host and network interface operating independently for transferring packets between protocol layers
US5448565A (en) * 1992-11-12 1995-09-05 International Business Machines Corp. Multiport LAN bridge
US5758075A (en) * 1994-07-29 1998-05-26 International Business Machines Corporation Multimedia communication apparatus and methods
US5764895A (en) * 1995-01-11 1998-06-09 Sony Corporation Method and apparatus for directing data packets in a local area network device having a plurality of ports interconnected by a high-speed communication bus
US5633865A (en) * 1995-03-31 1997-05-27 Netvantage Apparatus for selectively transferring data packets between local area networks
US5878028A (en) * 1996-06-06 1999-03-02 Advanced Micro Devices, Inc. Data structure to support multiple transmit packets for high performance

Cited By (169)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6738384B1 (en) 1997-09-17 2004-05-18 Sony Corporation Technique for optimizing cut-through for broadcast and multi-cast packets in a multi-port bridge for a local area network
US6816490B1 (en) 1997-09-17 2004-11-09 Sony Corporation Statistical learning technique in a multi-port bridge for a local area network
US6308218B1 (en) 1997-09-17 2001-10-23 Sony Corporation Address look-up mechanism in a multi-port bridge for a local area network
US6617879B1 (en) 1997-09-17 2003-09-09 Sony Corporation Transparently partitioned communication bus for multi-port bridge for a local area network
US6744728B1 (en) 1997-09-17 2004-06-01 Sony Corporation & Sony Electronics, Inc. Data pipeline timing optimization technique in a multi-port bridge for a local area network
US6363067B1 (en) 1997-09-17 2002-03-26 Sony Corporation Staged partitioned communication bus for a multi-port bridge for a local area network
US6252879B1 (en) 1997-09-17 2001-06-26 Sony Corporation Single counter for controlling multiple finite state machines in a multi-port bridge for local area network
US6442168B1 (en) 1997-09-17 2002-08-27 Sony Corporation High speed bus structure in a multi-port bridge for a local area network
US6446173B1 (en) 1997-09-17 2002-09-03 Sony Corporation Memory controller in a multi-port bridge for a local area network
US6751225B1 (en) 1997-09-17 2004-06-15 Sony Corporation Port within a multi-port bridge including a buffer for storing routing information for data packets received in the port
US6301256B1 (en) 1997-09-17 2001-10-09 Sony Corporation Selection technique for preventing a source port from becoming a destination port in a multi-port bridge for a local area network
US6341313B1 (en) * 1998-02-02 2002-01-22 Nec Corporation Flow controlling method and apparatus for network between processors
US6401146B1 (en) * 1998-02-06 2002-06-04 Lg Semicon Co., Ltd. Device and method for controlling PCI ethernet
US6577625B1 (en) * 1998-06-22 2003-06-10 Winbond Electronics Corp. Ethernet switch with a share memory structure and method for sharing memory
US7899052B1 (en) * 1999-01-27 2011-03-01 Broadcom Corporation Memory structure for resolving addresses in a packet-based network switch
US20060077995A1 (en) * 1999-01-27 2006-04-13 Broadcom Corporation Apparatus for ethernet PHY/MAC communication
US8072995B2 (en) 1999-01-27 2011-12-06 Broadcom Corporation Apparatus for ethernet PHY/MAC communication
US6601150B1 (en) * 1999-05-10 2003-07-29 Nortel Networks Limited Memory management technique for maintaining packet order in a packet processing system
US6781986B1 (en) * 1999-06-25 2004-08-24 Nortel Networks Limited Scalable high capacity switch architecture method, apparatus and system
US6584517B1 (en) * 1999-07-02 2003-06-24 Cypress Semiconductor Corp. Circuit and method for supporting multicast/broadcast operations in multi-queue storage devices
US20070086480A1 (en) * 1999-07-30 2007-04-19 Uri Elzur Associating a packet with a flow
US20040073778A1 (en) * 1999-08-31 2004-04-15 Adiletta Matthew J. Parallel processor architecture
US8316191B2 (en) 1999-08-31 2012-11-20 Intel Corporation Memory controllers for processor having multiple programmable units
US6775283B1 (en) * 1999-11-16 2004-08-10 Advanced Micro Devices, Inc. Passing vlan information through descriptors
USRE41849E1 (en) 1999-12-22 2010-10-19 Intel Corporation Parallel multi-threaded processing
US8738886B2 (en) 1999-12-27 2014-05-27 Intel Corporation Memory mapping in a processor having multiple programmable units
US9830285B2 (en) 1999-12-27 2017-11-28 Intel Corporation Memory mapping in a processor having multiple programmable units
US9824038B2 (en) 1999-12-27 2017-11-21 Intel Corporation Memory mapping in a processor having multiple programmable units
US9128818B2 (en) 1999-12-27 2015-09-08 Intel Corporation Memory mapping in a processor having multiple programmable units
US9824037B2 (en) 1999-12-27 2017-11-21 Intel Corporation Memory mapping in a processor having multiple programmable units
US9830284B2 (en) 1999-12-27 2017-11-28 Intel Corporation Memory mapping in a processor having multiple programmable units
US7751402B2 (en) 1999-12-29 2010-07-06 Intel Corporation Method and apparatus for gigabit packet assignment for multithreaded packet processing
US7480706B1 (en) * 1999-12-30 2009-01-20 Intel Corporation Multi-threaded round-robin receive for fast network port
US6952824B1 (en) 1999-12-30 2005-10-04 Intel Corporation Multi-threaded sequenced receive for fast network port stream of packets
US20060156303A1 (en) * 1999-12-30 2006-07-13 Hooper Donald F Multi-threaded sequenced receive for fast network port stream of packets
US6748445B1 (en) * 2000-02-01 2004-06-08 Microsoft Corporation System and method for exchanging data
US6975626B1 (en) 2000-03-31 2005-12-13 Sun Microsystems, Inc. Switched network for low latency communication
US7020161B1 (en) 2000-03-31 2006-03-28 Sun Microsystems, Inc. Prescheduling arbitrated resources
US7065580B1 (en) 2000-03-31 2006-06-20 Sun Microsystems, Inc. Method and apparatus for a pipelined network
US7061929B1 (en) * 2000-03-31 2006-06-13 Sun Microsystems, Inc. Data network with independent transmission channels
US6654811B1 (en) * 2000-04-13 2003-11-25 Nokia Inc. Backpressure arrangement in client-server environment
US6724759B1 (en) 2000-08-11 2004-04-20 Paion Company, Limited System, method and article of manufacture for transferring a packet from a port controller to a switch fabric in a switch fabric chipset system
US6804731B1 (en) 2000-08-11 2004-10-12 Paion Company, Limited System, method and article of manufacture for storing an incoming datagram in switch matrix in a switch fabric chipset system
US20040095927A1 (en) * 2000-08-11 2004-05-20 You-Sung Chang System, method and article of manufacture for updating a switching table in a switch fabric chipset system
US6731631B1 (en) 2000-08-11 2004-05-04 Paion Company, Limited System, method and article of manufacture for updating a switching table in a switch fabric chipset system
WO2002015496A1 (en) * 2000-08-11 2002-02-21 Paion Company, Limited Novel switch fabric chipset system and method
US7401126B2 (en) * 2001-03-23 2008-07-15 Neteffect, Inc. Transaction switch and network interface adapter incorporating same
US20020172195A1 (en) * 2001-03-23 2002-11-21 Pekkala Richard E. Apparatus amd method for disparate fabric data and transaction buffering within infiniband device
US20030110271A1 (en) * 2001-08-31 2003-06-12 Ramkumar Jayam Systems and methods for high speed data transmission using TCP/IP
US6760769B2 (en) * 2001-08-31 2004-07-06 Adaptec, Inc. Apparatus and methods for transmitting data at high speed using TCP/IP
US20030115338A1 (en) * 2001-08-31 2003-06-19 Ramkumar Jayam Apparatus and methods for receiving data at high speed using TCP/IP
US6981014B2 (en) * 2001-08-31 2005-12-27 Adaptec Corporation Systems and methods for high speed data transmission using TCP/IP
US7096247B2 (en) * 2001-08-31 2006-08-22 Adaptec, Inc. Apparatus and methods for receiving data at high speed using TCP/IP
US20030115337A1 (en) * 2001-08-31 2003-06-19 Ramkumar Jayam Apparatus and methods for transmitting data at high speed using TCP/IP
US20030067934A1 (en) * 2001-09-28 2003-04-10 Hooper Donald F. Multiprotocol decapsulation/encapsulation control structure and packet protocol conversion method
US20030120789A1 (en) * 2001-10-22 2003-06-26 Neil Hepworth Real time control protocol session matching
US7457862B2 (en) * 2001-10-22 2008-11-25 Avaya, Inc. Real time control protocol session matching
US20050216710A1 (en) * 2002-01-17 2005-09-29 Wilkinson Hugh M Iii Parallel processor with functional pipeline providing programming engines by supporting multiple contexts and critical section
US20030145173A1 (en) * 2002-01-25 2003-07-31 Wilkinson Hugh M. Context pipelines
US7181594B2 (en) * 2002-01-25 2007-02-20 Intel Corporation Context pipelines
US20050216822A1 (en) * 2002-02-18 2005-09-29 Hiroshi Kyusojin Data communication system, data communication management device and method, and computer program
US20030156597A1 (en) * 2002-02-21 2003-08-21 Sun Microsystems, Inc. Method and apparatus for speculative arbitration
US7352741B2 (en) 2002-02-21 2008-04-01 Sun Microsystems, Inc. Method and apparatus for speculative arbitration
US20100215046A1 (en) * 2002-03-22 2010-08-26 Juniper Networks, Inc. Logical separation and accessing of descriptor memories
US7916632B1 (en) 2002-03-22 2011-03-29 Juniper Networks, Inc. Systems and methods for handling packet fragmentation
US7158520B1 (en) 2002-03-22 2007-01-02 Juniper Networks, Inc. Mailbox registers for synchronizing header processing execution
US7616562B1 (en) 2002-03-22 2009-11-10 Juniper Networks, Inc. Systems and methods for handling packet fragmentation
US7212530B1 (en) 2002-03-22 2007-05-01 Juniper Networks, Inc. Optimized buffer loading for packet header processing
US7215662B1 (en) 2002-03-22 2007-05-08 Juniper Networks, Inc. Logical separation and accessing of descriptor memories
US7936758B2 (en) 2002-03-22 2011-05-03 Juniper Networks, Inc. Logical separation and accessing of descriptor memories
US7236501B1 (en) 2002-03-22 2007-06-26 Juniper Networks, Inc. Systems and methods for handling packet fragmentation
US7239630B1 (en) * 2002-03-22 2007-07-03 Juniper Networks, Inc. Dedicated processing resources for packet header generation
US7773599B1 (en) 2002-03-22 2010-08-10 Juniper Networks, Inc. Packet fragment handling
US20070183425A1 (en) * 2002-03-22 2007-08-09 Juniper Networks, Inc. Logical separation and accessing of descriptor memories
US7680116B1 (en) 2002-03-22 2010-03-16 Juniper Networks, Inc. Optimized buffer loading for packet header processing
US8085780B1 (en) 2002-03-22 2011-12-27 Juniper Networks, Inc. Optimized buffer loading for packet header processing
US7180893B1 (en) 2002-03-22 2007-02-20 Juniper Networks, Inc. Parallel layer 2 and layer 3 processing components in a network router
US20110142070A1 (en) * 2002-03-22 2011-06-16 Juniper Networks, Inc. Systems and methods for handling packet fragmentation
US7782857B2 (en) 2002-03-22 2010-08-24 Juniper Networks, Inc. Logical separation and accessing of descriptor memories
US7283528B1 (en) 2002-03-22 2007-10-16 Raymond Marcelino Manese Lim On the fly header checksum processing using dedicated logic
US20030223431A1 (en) * 2002-04-11 2003-12-04 Chavez David L. Emergency bandwidth allocation with an RSVP-like protocol
US7489687B2 (en) 2002-04-11 2009-02-10 Avaya. Inc. Emergency bandwidth allocation with an RSVP-like protocol
US8144733B2 (en) * 2002-05-03 2012-03-27 Intellectual Ventures I Llc Partitioned medium access control implementation
US20100002717A1 (en) * 2002-05-03 2010-01-07 Conexant, Inc. Partitioned Medium Access Control Implementation
US6701393B1 (en) * 2002-06-27 2004-03-02 Emc Corporation Systems and methods for managing storage location descriptors
US8593959B2 (en) 2002-09-30 2013-11-26 Avaya Inc. VoIP endpoint call admission
US20080151921A1 (en) * 2002-09-30 2008-06-26 Avaya Technology Llc Packet prioritization and associated bandwidth and buffer management techniques for audio over ip
US7877501B2 (en) 2002-09-30 2011-01-25 Avaya Inc. Packet prioritization and associated bandwidth and buffer management techniques for audio over IP
US8176154B2 (en) 2002-09-30 2012-05-08 Avaya Inc. Instantaneous user initiation voice quality feedback
US7877500B2 (en) 2002-09-30 2011-01-25 Avaya Inc. Packet prioritization and associated bandwidth and buffer management techniques for audio over IP
US8370515B2 (en) 2002-09-30 2013-02-05 Avaya Inc. Packet prioritization and associated bandwidth and buffer management techniques for audio over IP
US20080151886A1 (en) * 2002-09-30 2008-06-26 Avaya Technology Llc Packet prioritization and associated bandwidth and buffer management techniques for audio over ip
US20070133403A1 (en) * 2002-09-30 2007-06-14 Avaya Technology Corp. Voip endpoint call admission
US7359979B2 (en) 2002-09-30 2008-04-15 Avaya Technology Corp. Packet prioritization and associated bandwidth and buffer management techniques for audio over IP
US8015309B2 (en) 2002-09-30 2011-09-06 Avaya Inc. Packet prioritization and associated bandwidth and buffer management techniques for audio over IP
US20040073690A1 (en) * 2002-09-30 2004-04-15 Neil Hepworth Voice over IP endpoint call admission
US20040073641A1 (en) * 2002-09-30 2004-04-15 Muneyb Minhazuddin Instantaneous user initiation voice quality feedback
US20040073692A1 (en) * 2002-09-30 2004-04-15 Gentle Christopher R. Packet prioritization and associated bandwidth and buffer management techniques for audio over IP
US20040148555A1 (en) * 2003-01-24 2004-07-29 Dennis Blackburn Apparatus and method for accommodating loss of signal
WO2004071065A2 (en) * 2003-01-24 2004-08-19 Wegener Communications, Inc. Apparatus and method for accommodating loss of signal
WO2004071065A3 (en) * 2003-01-24 2006-08-17 Wegener Communications Inc Apparatus and method for accommodating loss of signal
US7263648B2 (en) * 2003-01-24 2007-08-28 Wegener Communications, Inc. Apparatus and method for accommodating loss of signal
US20050144338A1 (en) * 2003-02-27 2005-06-30 Fujitsu Limited Data transfer apparatus
US8396063B1 (en) * 2003-05-30 2013-03-12 Redpine Signals, Inc. Multi-thread media access controller with per-thread MAC address
USRE41919E1 (en) 2003-06-25 2010-11-09 Steve Olivier Rapid decryption of data by key synchronization and indexing
US20050063407A1 (en) * 2003-09-23 2005-03-24 Samsung Electronics Co., Ltd. Apparatus and method for maintaining high-speed forwarding tables in a massively parallel router
US7114016B2 (en) * 2003-12-23 2006-09-26 Intel Corporation Page-aware descriptor management
US20050138212A1 (en) * 2003-12-23 2005-06-23 Intel Corporation Page-aware descriptor management
US7447223B2 (en) * 2004-01-28 2008-11-04 Hewlett-Packard Development Company, L.P. Switching mesh with broadcast path redundancy
US20050163137A1 (en) * 2004-01-28 2005-07-28 Wakumoto Shaun K. Switching mesh with broadcast path redundancy
US20050216616A1 (en) * 2004-03-23 2005-09-29 Intel Corporation Inbound packet placement in host memory
US7284075B2 (en) * 2004-03-23 2007-10-16 Intel Corporation Inbound packet placement in host memory
US20050226147A1 (en) * 2004-04-06 2005-10-13 Oki Electric Industry Co., Ltd. Relay device
US7397816B2 (en) * 2004-04-06 2008-07-08 Oki Electric Industry Co., Ltd. Relay device
US7978827B1 (en) 2004-06-30 2011-07-12 Avaya Inc. Automatic configuration of call handling based on end-user needs and characteristics
US20060095611A1 (en) * 2004-11-02 2006-05-04 Standard Microsystems Corporation Hardware supported peripheral component memory alignment method
US8190796B2 (en) 2004-11-02 2012-05-29 Standard Microsystems Corporation Hardware supported peripheral component memory alignment method
US8649395B2 (en) * 2005-03-24 2014-02-11 Ixia Protocol stack using shared memory
US20120143979A1 (en) * 2005-03-24 2012-06-07 Olderdissen Jan Ra Protocol stack using shared memory
US20060230119A1 (en) * 2005-04-08 2006-10-12 Neteffect, Inc. Apparatus and method for packet transmission over a high speed network supporting remote direct memory access operations
US8458280B2 (en) 2005-04-08 2013-06-04 Intel-Ne, Inc. Apparatus and method for packet transmission over a high speed network supporting remote direct memory access operations
US20070165672A1 (en) * 2006-01-19 2007-07-19 Neteffect, Inc. Apparatus and method for stateless CRC calculation
US7889762B2 (en) 2006-01-19 2011-02-15 Intel-Ne, Inc. Apparatus and method for in-line insertion and removal of markers
US7782905B2 (en) 2006-01-19 2010-08-24 Intel-Ne, Inc. Apparatus and method for stateless CRC calculation
US20110099243A1 (en) * 2006-01-19 2011-04-28 Keels Kenneth G Apparatus and method for in-line insertion and removal of markers
US9276993B2 (en) 2006-01-19 2016-03-01 Intel-Ne, Inc. Apparatus and method for in-line insertion and removal of markers
US8699521B2 (en) 2006-01-19 2014-04-15 Intel-Ne, Inc. Apparatus and method for in-line insertion and removal of markers
US20080043750A1 (en) * 2006-01-19 2008-02-21 Neteffect, Inc. Apparatus and method for in-line insertion and removal of markers
US20070198720A1 (en) * 2006-02-17 2007-08-23 Neteffect, Inc. Method and apparatus for a interfacing device drivers to a single multi-function adapter
US8271694B2 (en) 2006-02-17 2012-09-18 Intel-Ne, Inc. Method and apparatus for using a single multi-function adapter with different operating systems
US20070208820A1 (en) * 2006-02-17 2007-09-06 Neteffect, Inc. Apparatus and method for out-of-order placement and in-order completion reporting of remote direct memory access operations
US20070226750A1 (en) * 2006-02-17 2007-09-27 Neteffect, Inc. Pipelined processing of RDMA-type network transactions
US8316156B2 (en) 2006-02-17 2012-11-20 Intel-Ne, Inc. Method and apparatus for interfacing device drivers to single multi-function adapter
US8489778B2 (en) 2006-02-17 2013-07-16 Intel-Ne, Inc. Method and apparatus for using a single multi-function adapter with different operating systems
US20070226386A1 (en) * 2006-02-17 2007-09-27 Neteffect, Inc. Method and apparatus for using a single multi-function adapter with different operating systems
US8078743B2 (en) 2006-02-17 2011-12-13 Intel-Ne, Inc. Pipelined processing of RDMA-type network transactions
US20100332694A1 (en) * 2006-02-17 2010-12-30 Sharp Robert O Method and apparatus for using a single multi-function adapter with different operating systems
US7849232B2 (en) 2006-02-17 2010-12-07 Intel-Ne, Inc. Method and apparatus for using a single multi-function adapter with different operating systems
US8032664B2 (en) 2006-02-17 2011-10-04 Intel-Ne, Inc. Method and apparatus for using a single multi-function adapter with different operating systems
US7617337B1 (en) 2007-02-06 2009-11-10 Avaya Inc. VoIP quality tradeoff system
US11829601B2 (en) 2008-02-28 2023-11-28 Memory Technologies Llc Extended utilization area for a memory device
US11550476B2 (en) 2008-02-28 2023-01-10 Memory Technologies Llc Extended utilization area for a memory device
US11182079B2 (en) 2008-02-28 2021-11-23 Memory Technologies Llc Extended utilization area for a memory device
US9367486B2 (en) 2008-02-28 2016-06-14 Memory Technologies Llc Extended utilization area for a memory device
US11494080B2 (en) 2008-02-28 2022-11-08 Memory Technologies Llc Extended utilization area for a memory device
US11907538B2 (en) 2008-02-28 2024-02-20 Memory Technologies Llc Extended utilization area for a memory device
US8218751B2 (en) 2008-09-29 2012-07-10 Avaya Inc. Method and apparatus for identifying and eliminating the source of background noise in multi-party teleconferences
US11775173B2 (en) 2009-06-04 2023-10-03 Memory Technologies Llc Apparatus and method to share host system RAM with mass storage memory RAM
US9208078B2 (en) 2009-06-04 2015-12-08 Memory Technologies Llc Apparatus and method to share host system RAM with mass storage memory RAM
US11733869B2 (en) 2009-06-04 2023-08-22 Memory Technologies Llc Apparatus and method to share host system RAM with mass storage memory RAM
US9983800B2 (en) 2009-06-04 2018-05-29 Memory Technologies Llc Apparatus and method to share host system RAM with mass storage memory RAM
US10983697B2 (en) 2009-06-04 2021-04-20 Memory Technologies Llc Apparatus and method to share host system RAM with mass storage memory RAM
US20120033673A1 (en) * 2010-08-06 2012-02-09 Deepak Goel Systems and methods for a para-vitualized driver in a multi-core virtual packet engine device
US20120137102A1 (en) * 2010-11-30 2012-05-31 Ramkumar Perumanam Consumer approach based memory buffer optimization for multimedia applications
US20130128739A1 (en) * 2011-11-21 2013-05-23 Broadcom Corporation Method and Device to Control Communication with Multiple Networks Based on Respective Quality of Service Requirements
US9417998B2 (en) * 2012-01-26 2016-08-16 Memory Technologies Llc Apparatus and method to provide cache move with non-volatile mass memory system
US11797180B2 (en) 2012-01-26 2023-10-24 Memory Technologies Llc Apparatus and method to provide cache move with non-volatile mass memory system
US20130198434A1 (en) * 2012-01-26 2013-08-01 Nokia Corporation Apparatus and Method to Provide Cache Move With Non-Volatile Mass Memory System
US10877665B2 (en) 2012-01-26 2020-12-29 Memory Technologies Llc Apparatus and method to provide cache move with non-volatile mass memory system
US10042586B2 (en) 2012-04-20 2018-08-07 Memory Technologies Llc Managing operational state data in memory module
US11226771B2 (en) 2012-04-20 2022-01-18 Memory Technologies Llc Managing operational state data in memory module
US9311226B2 (en) 2012-04-20 2016-04-12 Memory Technologies Llc Managing operational state data of a memory module using host memory in association with state change
US11782647B2 (en) 2012-04-20 2023-10-10 Memory Technologies Llc Managing operational state data in memory module
US9143562B2 (en) * 2012-04-27 2015-09-22 Hewlett-Packard Development Company, L.P. Managing transfer of data from a source to a destination machine cluster
US20130290471A1 (en) * 2012-04-27 2013-10-31 Rajat Venkatesh Managing transfer of data from a source to a destination machine cluster
US9164804B2 (en) 2012-06-20 2015-10-20 Memory Technologies Llc Virtual memory module
US9350658B2 (en) 2013-12-09 2016-05-24 International Business Machines Corporation FIFO affinity for multiple ports
US9166914B2 (en) * 2013-12-09 2015-10-20 International Business Machines Corporation FIFO affinity for multiple ports
US20150163141A1 (en) * 2013-12-09 2015-06-11 International Business Machines Corporation FIFO Affinity for Multiple Ports

Also Published As

Publication number Publication date
EP1086552A4 (en) 2001-10-04
AU4435699A (en) 1999-12-30
EP1086552A1 (en) 2001-03-28
CA2328220C (en) 2003-12-23
WO1999065193A1 (en) 1999-12-16
CA2328220A1 (en) 1999-12-16
DE69941697D1 (en) 2010-01-07
AU767085B2 (en) 2003-10-30
EP1086552B1 (en) 2009-11-25

Similar Documents

Publication Publication Date Title
US6067300A (en) Method and apparatus for optimizing the transfer of data packets between local area networks
US5884040A (en) Per-packet jamming in a multi-port bridge for a local area network
US6795886B1 (en) Interconnect switch method and apparatus
US7260104B2 (en) Deferred queuing in a buffered switch
US6192028B1 (en) Method and apparatus providing programmable thresholds for half-duplex flow control in a network switch
US6922408B2 (en) Packet communication buffering with dynamic flow control
US6526451B2 (en) Method and network device for creating circular queue structures in shared memory
EP1421739B1 (en) Transmitting multicast data packets
US7782849B2 (en) Data switch and switch fabric
US6091707A (en) Methods and apparatus for preventing under-flow conditions in a multiple-port switching device
US6754222B1 (en) Packet switching apparatus and method in data network
EP1014649A2 (en) Method and system of data transfer control
US6301256B1 (en) Selection technique for preventing a source port from becoming a destination port in a multi-port bridge for a local area network
US6601116B1 (en) Network switch having descriptor cache and method thereof
US6442168B1 (en) High speed bus structure in a multi-port bridge for a local area network
US6157951A (en) Dual priority chains for data-communication ports in a multi-port bridge for a local area network
US7113516B1 (en) Transmit buffer with dynamic size queues
US5748633A (en) Method and apparatus for the concurrent reception and transmission of packets in a communications internetworking device
US20040081094A1 (en) Method and system for output flow control in network multiplexers
US6256313B1 (en) Triplet architecture in a multi-port bridge for a local area network
EP1065835B1 (en) Packet memory management (PACMAN) scheme
US6810044B1 (en) Order broadcast management (IOBMAN) scheme
EP1065836B1 (en) Broadcast management scheme
JP2773198B2 (en) Data transfer device

Legal Events

Date Code Title Description
AS Assignment

Owner name: NETVANTAGE, CALIFORNIA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:BAUMERT, ROBERT J.;SEAMAN, ANTHONY W.;STAVES, SHERRE M.;REEL/FRAME:009247/0107

Effective date: 19980609

AS Assignment

Owner name: CABLETRON SYSTEMS, INC., NEW HAMPSHIRE

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:NETVANTAGE;REEL/FRAME:010388/0502

Effective date: 19990806

STCF Information on status: patent grant

Free format text: PATENTED CASE

AS Assignment

Owner name: ENTERASYS NETWORKS, INC., NEW HAMPSHIRE

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:CABLETRON SYSTEMS, INC.;REEL/FRAME:011219/0376

Effective date: 20000929

FEPP Fee payment procedure

Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

FPAY Fee payment

Year of fee payment: 4

AS Assignment

Owner name: OBSIDIAN, LLC, CALIFORNIA

Free format text: SECURITY AGREEMENT;ASSIGNOR:ENTERASYS NETWORKS, INC.;REEL/FRAME:017656/0552

Effective date: 20060516

Owner name: WELLS FARGO FOOTHILL, INC., CALIFORNIA

Free format text: SECURITY AGREEMENT;ASSIGNOR:ENTERASYS NETWORKS, INC.;REEL/FRAME:017656/0552

Effective date: 20060516

FEPP Fee payment procedure

Free format text: PAT HOLDER NO LONGER CLAIMS SMALL ENTITY STATUS, ENTITY STATUS SET TO UNDISCOUNTED (ORIGINAL EVENT CODE: STOL); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

FPAY Fee payment

Year of fee payment: 8

SULP Surcharge for late payment
AS Assignment

Owner name: WELLS FARGO TRUST CORPORATION LIMITED, AS SECURITY

Free format text: GRANT OF SECURITY INTEREST IN U.S. PATENTS;ASSIGNOR:ENTERASYS NETWORKS INC.;REEL/FRAME:025339/0875

Effective date: 20101109

AS Assignment

Owner name: ENTERASYS NETWORKS, INC., MASSACHUSETTS

Free format text: RELEASE AND REASSIGNMENT OF PATENTS AND PATENT APPLICATIONS AT REEL/FRAME NO. 17656/0552;ASSIGNORS:WELLS FARGO CAPITAL FINANCE, INC. (FORMERLY KNOWN AS WELLS FARGO FOOTHILL, INC.);ENTERPRISE COMMUNICATIONS FUNDING GMBH, AS SUCCESSOR IN INTEREST TO OBSIDIAN, LLC;REEL/FRAME:025406/0769

Effective date: 20101110

FPAY Fee payment

Year of fee payment: 12

AS Assignment

Owner name: ENTERASYS NETWORKS INC., MASSACHUSETTS

Free format text: TERMINATION AND RELEASE OF SECURITY INTEREST IN PATENTS AT REEL/FRAME NO. 25339/0875;ASSIGNOR:WELLS FARGO TRUST CORPORATION LIMITED;REEL/FRAME:031558/0677

Effective date: 20131031

AS Assignment

Owner name: SILICON VALLEY BANK, CALIFORNIA

Free format text: SECURITY AGREEMENT;ASSIGNOR:ENTERASYS NETWORKS, INC.;REEL/FRAME:036189/0509

Effective date: 20150724

AS Assignment

Owner name: EXTREME NETWORKS, INC., NEW HAMPSHIRE

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:ENTERASYS NETWORKS, INC.;REEL/FRAME:036467/0566

Effective date: 20150820

AS Assignment

Owner name: EXTREME NETWORKS, INC., CALIFORNIA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:ENTERASYS NETWORKS, INC.;REEL/FRAME:036538/0011

Effective date: 20150820

AS Assignment

Owner name: SILICON VALLEY BANK, CALIFORNIA

Free format text: AMENDED AND RESTATED PATENT AND TRADEMARK SECURITY AGREEMENT;ASSIGNOR:EXTREME NETWORKS, INC.;REEL/FRAME:040521/0762

Effective date: 20161028

AS Assignment

Owner name: SILICON VALLEY BANK, CALIFORNIA

Free format text: SECOND AMENDED AND RESTATED PATENT AND TRADEMARK SECURITY AGREEMENT;ASSIGNOR:EXTREME NETWORKS, INC.;REEL/FRAME:043200/0614

Effective date: 20170714

AS Assignment

Owner name: SILICON VALLEY BANK, CALIFORNIA

Free format text: THIRD AMENDED AND RESTATED PATENT AND TRADEMARK SECURITY AGREEMENT;ASSIGNOR:EXTREME NETWORKS, INC.;REEL/FRAME:044639/0300

Effective date: 20171027

AS Assignment

Owner name: ENTERASYS NETWORKS, INC., CALIFORNIA

Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:SILICON VALLEY BANK;REEL/FRAME:046047/0223

Effective date: 20180501

Owner name: BANK OF MONTREAL, NEW YORK

Free format text: SECURITY INTEREST;ASSIGNOR:EXTREME NETWORKS, INC.;REEL/FRAME:046050/0546

Effective date: 20180501

Owner name: EXTREME NETWORKS, INC., CALIFORNIA

Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:SILICON VALLEY BANK;REEL/FRAME:046051/0775

Effective date: 20180501